Adds the portal workspace: /my/visit/new starts a visit; /my/visit/<id> shows the
add-as-you-go workspace (add buttons -> existing forms carrying ?visit_id, a
deferred client+funding form, and a Complete button). Accessibility forms launched
from a visit save as a DRAFT linked to it (JS carries visit_id into the form; the
controller captures it and skips the per-assessment SO) - the VISIT completion then
creates the grouped per-funding sale orders.
NOT YET: express/ADP form visit-linking, email consolidation, polished tablet UI.
Untested locally (Enterprise dep) - clone verification pending.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The visit groups its ADP assessments by funding type onto ONE ADP order (first
device creates the SO via the existing express completion; the rest attach),
enforcing the combination rule: at most one seated-mobility device (manual WC /
power WC / scooter) + optionally one walker, no duplicates. Also fixes a Phase 1b
bug - it called action_complete() (needs signatures, returns an action dict) for
ADP; now uses action_complete_express() which returns the SO.
Untested locally (Enterprise dep) - clone verification pending.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Adds 'scooter' as a 4th ADP equipment type with a lean Express-form section
(scooter type + max range) and the power-mobility home-accessibility hard rule
(scooter + powerchair): "is the home usable inside and outside, no lifting?" - if No,
prompts adding an accessibility item (ramp / porch lift). Captures
x_fc_power_home_accessible + notes; the section toggles via the existing
equipment-select JS; controller parses the new fields.
NOT YET: ADP multi-device + combination rules (the bigger restructure).
Untested locally (Enterprise dep) - to be verified on a westin-v19 clone.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Adds fusion.assessment.visit: the hub that bundles a home visit's assessments and,
on completion, groups its ACCESSIBILITY assessments by funding workflow
(x_fc_sale_type) and creates ONE draft sale order per workflow, reusing the existing
MOD/ODSP/etc. pipelines + the chatter-note pattern. ADP assessments keep one-SO-each
for now (ADP multi-device grouping is Phase 2).
- New model + sequence (VISIT/YYYY/NNNN) + ACL + backend list/form/menu.
- visit_id added to fusion.assessment, fusion.accessibility.assessment, sale.order.
- action_complete_visit() group-and-routes; MOD $15k cap + income-threshold flag
surfaced (informational, no auto-enforcement).
NOT YET: completion-email consolidation; ADP multi-device (Phase 2); portal
add-as-you-go workspace (Phase 3). Untested locally (Enterprise knowledge dep) -
to be verified on a westin-v19 clone before prod.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Coerce an unexpected/tampered funding_source to direct_private instead of passing
it raw into create() (which would raise on the Selection field). Mirrors the
/book-assessment controller; the whitelist is derived from the model selection so
it auto-covers hardship and any future values.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Reps can now mark an accessibility assessment's funding source on the web form
(Private / March of Dimes / ODSP / WSIB / Hardship / Insurance / Other) so the
generated draft sale order routes to the correct funding pipeline instead of
always defaulting to private pay. Adds Hardship to the x_fc_funding_source
selection + sale_type_map; the new form <select> is auto-serialised by the
existing FormData submit, and accessibility_assessment_save now maps
funding_source -> x_fc_funding_source. The model + SO routing were already in
place (2026-04 audit fix) — this closes the form + controller gap.
Plan: docs/superpowers/plans/2026-06-02-accessibility-funding-selector.md
Spec: docs/superpowers/specs/2026-06-02-assessment-visit-funding-design.md
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Brainstormed design: bundle a home visit's assessments (ADP + accessibility),
measurement-first with client/funding deferred, add-as-you-go workspace,
per-item funding selector (fixes the March-of-Dimes routing gap), and on
completion group items into ONE draft sale order per funding workflow
(ADP / MOD / ODSP / Hardship / private) reusing the existing pipelines.
Adds ADP multi-device + combination rules, a new mobility-scooter type, and a
power-mobility home-accessibility rule that feeds the accessibility upsell.
v1 keeps manual quotation (no auto-pricing); MOD $15k cap is a reminder only.
Phased 1-3; risks + file map included.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Rename module fusion_authorizer_portal -> fusion_portal everywhere:
manifest/assets, controllers, models, views, JS (odoo.define + asset URLs),
migration MODULE constants; plus cross-module refs in fusion_schedule,
fusion_repairs, fusion_quotations (depends + inherit_id) and the pdf_filler
import in fusion_claims. Add rename_module.sql for the one-time in-place DB
rename (ir_module_module, ir_model_data, ir_ui_view.key,
ir_module_module_dependency) required on installed envs before -u fusion_portal.
Document the rename gotcha as rule 16 in CLAUDE.md.
Redesign the Accessibility Assessment selector: replace Font Awesome icon tiles
with photo-banner cards using 7 optimized images (1000x750 PNG -> 800x600 JPEG,
~8MB -> 488KB), per-type colour accent bar + centered pill button, hover
lift/zoom. Images ship as module static files so they deploy/sync with the module.
Drop the regenerable graphify-out cache from the module.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Recomputing only x_fclk_break_minutes left historical x_fclk_net_hours / x_fclk_overtime_hours stale (add_to_compute+flush of one field does not cascade to dependents). Recompute the full chain in dependency order. Caught verifying the entech deploy.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Statutory unpaid break now deducts automatically from worked hours on every path - portal, kiosk, NFC, auto-clock-out cron, AND manual backend entry.
- new fusion.clock.break.rule per-province table (seed Ontario 5h->30, 10h->+30), resolved from the employee's company province with a global default fallback
- x_fclk_break_minutes is now a single idempotent stored compute (statutory(worked_hours) + penalties), replacing the 4 duplicated write sites (_apply_break_deduction x3 callsites + auto-clock-out cron + penalty write)
- retire break_threshold_hours (superseded by per-rule break1_after_hours); post-migrate drops the param and recomputes historical breaks
- 11 tests all green; module install + 19.0.4.1.0 migration verified on modsdev
Bump 19.0.4.0.3 -> 19.0.4.1.0.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
People aren't good with 24h. Default Clock-In/Out are now AM/PM dropdowns (15-min
grid) instead of 24h float_time inputs. Stored value stays the float-string
(e.g. '9.0'), so all downstream float(get_param(...)) reads are unchanged;
persisted manually with get-snap for any off-grid value. Bump 19.0.4.0.3.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Settings tidy-up: one 'Kiosk' block holding both PIN Kiosk and NFC Kiosk
(clearly described so users know which is which), each with an Open-kiosk
button when enabled; Corrections + Sounds split into a 'Portal' block. Move
Auto-Wipe Photos under Photo Verification (was hidden for PIN-only clients).
Bump 19.0.4.0.0 -> 19.0.4.0.1.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
A proper shared-device PIN kiosk for clients who don't want NFC: photo-tile grid
(+search) -> tap -> PIN (or first-use create) -> optional master-gated selfie ->
clock, in the NFC kiosk's dark glass + brand-gradient style. Built as an Odoo 19
Interaction; new pin_kiosk.scss (scoped); reworked clock_kiosk.py
(search +avatar/has_pin, verify_pin needs_setup, set_pin, clock via kiosk location).
Drops the redundant kiosk_pin_required (PIN always required); relabels the company
kiosk location; adds a PIN-kiosk app icon. Opt-in via enable_kiosk (off by default).
HttpCase tests added. Bump 19.0.4.0.0.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The global enable_photo_verification toggle only fed the portal get_settings flag;
the actual writes ignored it — the NFC kiosk gated on nfc_photo_required and the
portal on location.require_photo, so photos were captured even with the toggle OFF.
Now it's the master: OFF => no photo captured/stored anywhere (NFC kiosk config +
tap, and portal check-in); ON => per-location / NFC settings apply. Test + help
text updated. Bump 3.16.0 -> 3.16.1.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Audit of all 41 settings found 3 that were shown but read nowhere, and 17 Boolean
toggles that couldn't be turned OFF.
- Remove grace_period_minutes (orphaned by the schedule-driven cron rewrite) and
weekly_overtime_threshold (never implemented): field + view + seed.
- enable_ip_fallback now actually gates _verify_location's IP-whitelist check
(default ON to preserve current behaviour).
- All 17 fusion_clock Boolean settings now persist explicitly as 'True'/'False'
via a _FCLK_BOOL_PARAMS loop in get_values/set_values (config_parameter Booleans
can't store False, so OFF never stuck). Add round-trip tests. Bump 3.15.2 -> 3.16.0.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The Pay Period Anchor Date was a free-text Char. Make it a fields.Date (date
picker) persisted manually in get_values/set_values as 'YYYY-MM-DD' under
fusion_clock.pay_period_start (res.config.settings Date fields don't round-trip
via config_parameter in Odoo 19). Reader code unchanged. Bump 3.15.1 -> 3.15.2.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
_resolve_tz fell back to env.company.tz, which raises AttributeError for any
user without a personal tz (surfaced by the new list-wide pay-period filters,
which resolve a company-level tz). Use env.company.partner_id.tz. Regression
test added. Bump 3.15.0 -> 3.15.1.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Reuse the existing Pay Period setting (Frequency + Anchor) as the single
source of truth via a shared pure helper (models/pay_period.py); fusion.clock.report
delegates to it. Add Current/Previous/Next Pay Period filters to the attendance
search view (search-method computed booleans on hr.attendance), a Bi-Weekly Period
picker wizard (pick start -> auto +2 weeks, editable; Apply opens the filtered list)
reachable from an Attendance menu item and a dashboard tile. Window follows the
configured frequency; TZ-correct via local-day boundaries. Bump 3.14.4 -> 3.15.0.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
hr_attendance's action is gantt-first and the native gantt timeline renders collapsed until a manual resize; open viewType:list so the button lands on a working list. Bump 3.14.3 -> 3.14.4.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Quick Actions were at the very bottom, so managers had to scroll past the whole team band to reach the nav shortcuts. Relocate the block to just above the Team/Org section (still below the personal band everyone has). Bump 3.14.2 -> 3.14.3.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Plain height:100%+overflow-y:auto did not scroll under the flex action container. Use the proven pattern: root flex column height:100%; inner .fclk-dash-wrap flex:1; min-height:0; overflow-y:auto. Bump 3.14.1 -> 3.14.2.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Drop the 1200px centred cap (wasted side space) and make .fclk-dash height:100%; overflow-y:auto so tall content scrolls. Bump 3.14.0 -> 3.14.1.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
- Rework /fusion_clock/dashboard_data into a personal block (everyone)
plus a team block (team lead = direct reports, manager = org-wide).
A regular employee's payload never contains another employee's data.
- New OWL stacked layout: gradient KPI cards (Today/Week/OT/Streak),
Today's Shift, Recent Activity, Upcoming Leave, Recent Penalties; team
band adds Present/Absent/Late/Pending, roster, and Needs Attention.
- Dark/light via compile-time $o-webclient-color-scheme branching;
drop the old runtime html.o_dark dashboard block.
- Open the Dashboard menu to group_fusion_clock_user (lead/manager imply).
- Add HttpCase permission/no-leak tests. Bump 3.13.2 -> 3.14.0.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
- Attendance now groups the operational records: All Attendances, Leave
Requests, Correction Requests, Penalties (Leaves + Penalties moved in from
top level).
- Scheduling groups all schedule-building: Shift Planner, Scheduled Shifts,
Shifts (templates, moved from Configuration), Schedule Audit.
- Configuration: Settings, Locations, Enroll NFC Card (the NFC wizard moved in
from top level).
- Removed the duplicate top-level Locations menu (kept the one under Config).
Only parent/sequence changed; no actions/views touched. Live on entech 3.13.2.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The "My Schedule" portal page read only published planning.slot (Odoo Planning),
but team leads post in the fusion_clock Shift Planner, which writes
fusion.clock.schedule -> so posted schedules never appeared. Merge both sources:
the page now lists published planning.slot AND posted fusion.clock.schedule
(employee, state=posted, not OFF, within the 60-day horizon), sorted together.
Verified on entech: Garry's 7 posted shifts (Jun 1-7) now render. 19.0.1.5.0.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
iOS Safari date inputs have a large intrinsic min-width that can break a flex
row; switch .fclk-leave-daterange to grid 1fr 1fr + min-width:0 on the inputs
so the two fields always share the row and shrink. Also changes the bundle hash
to force iOS to drop the cached CSS. Live on entech 19.0.3.13.1.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Request Leave now takes a From/To date range instead of a single day (the To
field is optional -> single-day). Added date_to to fusion.clock.leave.request
(start kept as leave_date), with overlap detection on submit and a date_to >=
leave_date constraint. The absence check and reports now treat a leave as
covering its whole span. The form shows two date inputs; the controller accepts
date_from/date_to (the old single leave_date payload is still honoured). A
migration backfills date_to = leave_date for existing rows.
Live and verified on entech 19.0.3.13.0.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The Schedule tab is injected into the Clock/Timesheets/Reports navs via xpath
inherits, but the two payslip templates (portal_payslip_list_page,
portal_payslip_detail_page) had no inherit, so Payslips showed only 4 tabs.
Add the matching inherits. Verified on the rendered /my/clock/payslips page:
5 nav items incl. Schedule. Live on entech 19.0.1.4.0.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Verified from the live DOM that fusion_plating_portal wraps the app in
#wrapwrap > main > .o_fp_portal_shell > .o_fp_portal_main > #wrap.o_portal_wrap
> .container. The white frame was .o_fp_portal_shell (+ .container max-width),
which my earlier wrapper-neutralisation didn't target. Add the shell + inner
main + force all wrappers transparent/full-width/no-padding under
body:has(.fclk-app). Live on entech 19.0.3.12.4.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
These round-2 portal fixes (white-border wrapper neutralisation in
portal_clock.css, and the Payslips nav tab on the fusion_planning Schedule
page) were briefly bundled into a concurrent NFC commit that a parallel session
then rebased, dropping them from main. They are deployed and verified on entech
(fusion_clock 3.12.3 / fusion_planning 1.3.0); re-committing so git matches.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Odoo 19 silently ignores the legacy `_sql_constraints` list (repo CLAUDE.md
rule 9), so it never created a DB constraint — two employees could be assigned
the same x_fclk_nfc_card_uid and the NFC tap's search(limit=1) then picked an
arbitrary one. Replace it with a declarative models.UniqueIndex carrying a
partial WHERE predicate, so uniqueness is enforced only when a UID is set;
employees without a card keep sharing a blank/NULL value.
Makes test_nfc_models.TestNfcModels.test_card_uid_is_unique_when_set pass.
Verified on entech (DB admin): 0 pre-existing duplicate UIDs, full upgrade +
61/61 fusion_clock tests green, and the unique partial index
hr_employee_fclk_nfc_card_uid_unique now exists.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
- White border on every portal page: the .fclk-app full-bleed relied on exact
negative margins to cancel the portal layout's container padding; when it
didn't match, the white page chrome showed through. Match the PAGE background
to the app (light #f3f4f6 / dark #0f1117, via body:has(.fclk-app)) so the
gutter is invisible, and clip horizontal overflow.
- Timesheets not responsive: the 6-column table crammed/wrapped on phones.
Replaced the table with stacked cards (date + net up top, in -> out, then
break / location / Correct) that read cleanly at any width. Correction-link
data attributes preserved; the xpath-inherited .fclk-nav-bar untouched.
Live on entech 19.0.3.12.2 (both rules verified in the served frontend bundle).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Internal staff now land on /my/clock with no customer sidebar; new
finalized-payslip portal under /my/clock/payslips (inline paystub from
payslip.line_ids + PDF). Customers' portal is unchanged. Live on entech.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
A shared portal_employee_navbar template broke fusion_planning, which
xpath-inherits each clock page's inline fclk-nav-bar to inject its
Schedule tab (anchored on a[@href='/my/clock/timesheets']). Revert to the
original inline-nav pattern on all four pages and append the Payslips item
to each — zero changes needed in fusion_planning.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Reminders, absence detection, late/early penalties, and auto-clock-out are now
driven by each employee's real schedule (posted planner entry -> recurring
shift), never the global 9-5 default. Employees who aren't scheduled get no
reminders/absence. Overtime past the scheduled end is never cut off — auto
clock-out only fires at a max-shift safety cap (default raised 12 -> 16h). Team
leads build the planner in draft and Post it (publishes + emails employees).
- hr.employee._get_fclk_day_plan: explicit `scheduled` flag; posted-only planner
entries (drafts ignored), else recurring shift covering that weekday, else
not-scheduled; sources 'schedule'/'shift'/'none'.
- fusion.clock.shift: day_mon..day_sun weekday pattern + covers_weekday().
- fusion.clock.schedule: draft/posted state + posted_date; planner edits reset
to draft; fclk_email_posted_week notification.
- Rewrote the reminder / absence / auto-clock-out crons: schedule-gated,
per-employee savepoints, OT-aware cap, weekend hardcode removed.
- Penalties + all three clock-in paths skip days the employee isn't scheduled.
- shift_planner: Post Week route + planner Post button + draft count.
- Migration backfills pre-existing schedule entries to 'posted' so they keep
driving automation after upgrade.
- Tests: resolver matrix, cron gating, OT cap; fixed the existing planner test
for the new state/source semantics.
Design: docs/superpowers/specs/2026-05-30-schedule-driven-attendance-design.md
Frontend footprint kept at zero to avoid colliding with the concurrent
employee-portal (payslips) work.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
entech runs the enterprise hr_payroll module (not the custom fusion_payroll),
whose hr.payslip lacks employee_cpp/ytd_* fields. Render the inline paystub
from payslip.line_ids (name + total) so it works on any payroll provider.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Odoo 19's _get_report() resolves a dotted string report_ref through
env.ref() as an XML ID, which lands on the QWeb view rather than the
ir.actions.report action. Pass the action id (matches every other
_render_qweb_pdf call site in the repo).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Separate internal employees from the customer portal: suppress the
fusion_plating_portal sidebar for internal users, redirect them to the
clock page, and add a finalized-payslip view (inline paystub + optional
PDF) under /my/clock/payslips in fusion_clock.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Posted-schedule/recurring-shift drives reminders, absence, penalties, and
auto-clock-out (never the global 9-5 default); overtime never cut (auto-close
only at a safety cap); team-lead draft->post workflow with employee notify.
Frontend footprint kept at zero to avoid colliding with the concurrent
fusion_plating employee-portal session.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The website module injects a fixed "frontend->backend" nav
(.o_frontend_to_backend_nav — the floating apps-grid/edit button) on every
frontend page for any internal user. Since the kiosk account is an internal
user, that button let a kiosk user tap through to the Odoo backend.
Hide it with a page-scoped inline style in the kiosk template head, so it's
suppressed only on /fusion_clock/kiosk/nfc and the real website keeps its nav.
Live as 19.0.3.11.8 (verified the rule is in the rendered template).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Privacy/space housekeeping for the kiosk verification selfies. A new daily cron
(_cron_fusion_wipe_old_photos) deletes the photo attachments on attendances
whose clock-in is older than fusion_clock.photo_retention_days (default 60).
Only the images are removed — attendance records, worked hours and penalties
are kept. Clearing the attachment-backed binary reclaims filestore space.
- Configurable in Settings → Fusion Clock → NFC Kiosk ("Auto-Wipe Photos After
(days)"); set 0 to disable.
- Wipes all three photo fields (NFC check-in/out + legacy portal photo),
batched with per-batch savepoints.
- tests/test_photo_retention.py covers wipe-old / keep-recent / retention=0.
Verified live on entech (19.0.3.11.7) via a rollback-only dry run: a 70-day
shift's photos were wiped (record + 8h hours preserved) while a 5-day shift's
photo was kept; nothing persisted. 0 attendances currently exceed 60 days.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The kiosk captures and stores a photo on every tap (x_fclk_check_in_photo /
x_fclk_check_out_photo on hr.attendance), but no view displayed those — the
form only showed the legacy portal field x_fclk_checkin_photo, so the NFC
photos were invisible in the UI. Add a "Verification Photos" group showing the
clock-in and clock-out photos (plus the legacy portal photo), each hidden when
empty. (The activity log has no image field — photos live on the attendance.)
Live as 19.0.3.11.6.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The result card showed x_fclk_net_hours = worked_hours − break − early-out
penalty minutes. Tapping out before the scheduled end adds a 15-min early-out
penalty to the break field, so short shifts clamped to 0 → "Worked 0h 0m".
Show GROSS attendance.worked_hours (the actual clock-in → clock-out elapsed
time) instead, and format adaptively (Xh Ym / Ym / Ys) so brief shifts and
quick tests don't all read 0. Net-of-deductions stays in the payroll reports.
Live as 19.0.3.11.5 (verified worked_hours computes correctly in the DB).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Root-caused on live entech (not guessed):
- The kiosk runs as a non-HR operator (uid 141) who gets AccessError reading
hr.employee images, so /web/image served a placeholder. Point the result-card
avatar at hr.employee.public/avatar_128 — verified readable as the operator,
returns the real photo. (Odoo's own UI uses .public for employee images.)
- The Odoo profile/preferences avatar is res.users → res.partner.image_1920,
which the capture never wrote. Propagate the captured photo to the linked
user's partner image so the profile updates too.
- Enlarge the capture oval (it was small): stage 62vh/520px, guide width 64%.
Live as 19.0.3.11.4. Also backfilled the existing test photo to the user's
partner image so the profile shows it without re-capturing.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
- Profile photo DID save (verified: image_1920 attachment persists); the
"doesn't update" was a browser-cache miss. Add ?unique=<write_date> to the
result-card avatar URL so a freshly-captured photo shows on clock in/out.
- Capture now starts a 10-second countdown (time to get into frame) then
auto-snaps; the button toggles to Cancel while counting.
- Face guide is now a VERTICAL oval (aspect-ratio 3/4) over a portrait stage —
it was rendering horizontal. Faces are taller than wide.
Deployed live to entech (LXC 111) as 19.0.3.11.3; frontend bundle verified to
compile clean and contain the new rules.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Odoo's Sass compiler evaluates the built-in min() function and errors with
"Incompatible units: 'px' and 'vw'" on `width: min(86vw, 380px)`, which broke
the entire web.assets_frontend bundle (kiosk + all portal pages unstyled).
Equivalent, compiler-safe: `width: 86vw; max-width: 380px;`.
Verified: forced a fresh frontend bundle compile on entech — no Incompatible
-units error, served CSS contains the compiled --pin rule. Live as 19.0.3.11.2.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The --pin panel used width:auto, so in the centred flex overlay it
collapsed to its content width and crushed the 3-column numpad. Give it
a definite width (min(86vw, 380px)) and make the keys proper tappable
squares (min-height 60px, 1.6rem font).
Deployed live to entech (LXC 111) as 19.0.3.11.1.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
NFC kiosk:
- Add "📷 Photo" action to every Manage-page employee row and to the
post-enroll result card, so a manager can set/replace a profile photo
at any time (previously only surfaced when the employee had no image).
- Slim the Manager PIN pad: dedicated --pin panel variant (max-width 360px,
reduced padding) with a tighter numpad, removing the oversized whitespace.
Deployed live to entech (LXC 111) as 19.0.3.11.0.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Kiosk work across this session (19.0.3.6.0 -> 19.0.3.10.0):
- Program-from-unknown-tap: amber prompt -> Manager PIN -> pick/create employee
-> binds the captured UID (no re-tap). Reassign moves a card between employees.
- Manager page (gear, when unlocked): search employees + tag status; assign/re-tag,
clear tag, archive employee, + new employee. Server-gated by the enroll password.
- Screen lock: kiosk starts locked (tap-only); Unlock -> Manager PIN, Lock button;
PIN remembered for the session so the gear never re-prompts.
- Sounds: pleasant + loud sine chimes (rising in / descending out) + a low "denied"
tone for wrong/unknown taps. Gated by fusion_clock.enable_sounds.
- Guided profile-photo capture for employees with no picture (clock-in or enroll):
live camera + oval face guide -> capture -> preview -> save to hr.employee.
- PIN no longer re-renders per digit; centered result card; 12h time; clock-out shows
"Worked Xh Ym this shift"; modern clock idle icon; faster animations/result timers;
session keep-alive so the kiosk login never expires.
- New endpoints: create_employee, clear_tag, delete_employee (archive), verify_pin,
save_profile_photo; enroll gains force-reassign.
- Docs: fusion_clock is now developed in Claude Code (dropped Cursor references).
Spec/plan under fusion_clock/docs/superpowers/. Deployed live on entech
(odoo-entech / LXC 111 on pve-worker5), v19.0.3.10.0.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
- PWA manifest on the NFC kiosk page so it installs as a full-screen
home-screen app (Chrome "Install" / Safari "Add to Home Screen").
- Dedicated "Kiosk Operator" permission + gated "Fusion Clock Kiosk"
top-level app (act_url -> /fusion_clock/kiosk/nfc). Kiosk controllers
accept Manager OR Kiosk Operator; all kiosk data ops already run sudo.
- Fix 403: read the company kiosk location via sudo on page-load and tap
(Kiosk Operator has no fusion.clock.location ACL).
- Odoo 19 permissions UX: ir.module.category + res.groups.privilege so
User/Team Lead/Manager and Kiosk Operator appear as application-access
dropdowns on the user form (no developer mode). Short group display names.
- Docs: note res.groups.privilege as the Odoo 19 category_id replacement.
Deployed live to entech (odoo-entech / LXC 111 on pve-worker5). v19.0.3.6.0.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Default-hide six order-line columns that aren't part of the plating
view (Product/product_template_id, Description Template, Specification,
Delivered Qty, Invoiced Qty, Taxes) by flipping them to optional="hide".
They stay available via the optional-columns toggle. Default-visible set
is now Customer-Facing, Part, Process/Recipe, Thickness, Serial, Job #,
Effective Deadline, Qty, Unit, Unit Price, Amount — for both quotations
and sales orders.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Remove the unused Customer Specification field and the redundant
Delivery Date (Lead Time covers it) from the customer-facing SO
confirmation and invoice PDFs (portrait + landscape). SO info row goes
5->4 columns (Delivery Date gone); the Customer Job # / Spec / Delivery
Method row goes 3->2 (Spec gone). Internal docs (traveller, sticker) and
the CoC process "Specification(s)" section are left untouched.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Tooling/additional charge lines (any product line with no part catalog)
no longer print in the parts table — they render in the totals block
under the subtotal with their entered label + amount. Subtotal is now
parts-only; tax + grand total are unchanged (the charge is still a real
taxed line in the data). Applies to SO confirmation and invoice, both
portrait and landscape. Also aligns the invoice S/N cell to the SO's
multi-serial rendering.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Add a cohesive, restrained colour layer using the existing express
tokens (light/dark aware): faint plum gradient washes on the PO card,
legend bar, table header, and Order Summary header; filled accent
gradient pills (EXPRESS / CAD); accent rules on the section title,
summary header, and Grand Total footer. Adds an $xpr-accent-tint token
plus four composed gradient tokens.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
The right-aligned value column squeezed the Additional Charge and Tax
dropdowns to a sliver. Move each picker into the (wider) label column,
stacked under its label at full width, so every value cell is now a
single amount that lines up cleanly in the right column under the
vertical divider.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Switch the summary rows from a flex space-between layout to a fixed
two-column grid (label | value) so a vertical divider on the label
cell's right edge lines up across every row. Values are right-aligned
into a clean amount column; the Grand Total footer keeps the divider at
the heavier rule weight.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Restyle the Order Summary card into a clean bordered table — a tinted
"Order Summary" caption bar, a divider line under every row, and an
accent-tinted Grand Total footer with a strong top rule. Uses the
existing light/dark express tokens so it renders correctly in both
colour schemes.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Express order entry now has a single "Lot Order" toggle on the header
instead of a per-line "Lot" checkbox. When on, every line shows Lot
Total and prices as a flat lot (unit price derived = lot total / qty,
qty preserved for production); when off, the Lot Total column is hidden
and lines price per unit as usual. Keeps the order summary clean for the
common per-unit case.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Bite-sized TDD plan: charge-type model + config UI, wizard charge/tax fields,
totals = one tax on (subtotal+charge), per-line lot pricing, SO-create tax on
all lines + typed charge line, and the express summary/line view changes.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Direct/Express order entry: a searchable/creatable fp.additional.charge.type
replaces the fixed Tooling Charge; one order-level account.tax applies to
(subtotal + charge); per-line lot pricing (flat lot total, derived unit price,
qty preserved). Reordered summary. Quotes out of scope.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Grant Odoo Billing (account.group_account_invoice) to group_fp_manager via
implied_ids; Quality Manager + Owner inherit it. Billing only (not Accountant);
the SO-origin workflow gate in fusion_plating_jobs is unchanged, so managers
invoice from the Sale Order's Create Invoice action. Tests assert Manager/Owner
get Billing and Shop Manager does not.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Read-only per-part version history (version#, reference, customer-facing,
order, by/when) below the curated templates list.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Wizard line (direct + express) and SO line now pre-fill BOTH internal +
customer-facing from the part's latest version (fallback to
default_specification_text), without clobbering typed text.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Each part-bearing line writes a deduped version (final order# + date) via
_fp_save_description_version, after the parent-number rename so the title
reflects the confirmed order number.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Bite-sized TDD plan: version model + part load/save helpers, save-on-confirm
hook, wizard + SO-line auto-load, and the part Descriptions-tab history list.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Dedicated fp.part.description.version model: latest auto-loads both internal +
customer-facing into a new order line; on SO confirm, a changed description
saves a new version titled "S#### · date". Browsable per-part history;
default_specification_text kept synced. SO surfaces only (not quotes).
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
The order-line onchange still auto-ticks "Save as Default" (so a new part's
spec is remembered next time) — only the explanatory popup is removed, per
client request. The ticked checkbox on the line is the cue now.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Carrier/service/weight inputs + Generate Label + Mark Shipped, shown when the
job is awaiting_ship and gated read-only ("Waiting on: WO-xxxx") until every
job on the order is ready. Reuses workspace card tokens; dark-mode accent
override included.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
generate_label (sudo'd FedEx machinery) and mark_shipped (as the technician),
both enforcing the order-level ship-together gate. /load now returns a
shipping block (carrier/service/weight + readiness + any existing label)
when the job is awaiting_ship.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
_fp_order_ship_state + _fp_mark_order_shipped enforce spec D4 ship-together:
the order ships only when every active job on it is awaiting_ship/done.
Shared by the tablet shipping endpoints and /fp/workspace/load.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Spec D5 delivery-completion set. Dispatch records (route/vehicle/pickup)
stay read-only for technicians.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
ACL: grant group_fp_technician write+create on fp.receiving / line / damage.
sudo the internal sale.order x_fc_receiving_status write so a non-privileged
technician isn't blocked inside action_mark_counted / action_close.
Tests deferred to entech (local Docker unavailable this session).
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Bite-sized TDD plan across receiving ACL+sudo, delivery ACL, fp.job
ship-readiness helpers, shipping endpoints, and the workspace shipping
panel. Also patches the spec to record the sale.order status-write sudo
fix found during planning.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Design for letting Technicians receive a confirmed order and ship a finished
order from the fp_job_workspace tablet surface. Receiving is ACL-only (the
panel + endpoints already exist); shipping adds a workspace panel + two
sudo-backed endpoints (generate label, mark shipped) gated on all order jobs
being awaiting_ship.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
The CoC body now renders English + the French translation together, so the
separate "Certificat de Conformité (Français)" print option was redundant.
- Removed the action_report_coc_fr report action and the now-dead
report_coc_fr template; renamed action_report_coc_en to "Certificate of
Conformance" (print filename "CoC - <name>").
- fp_notification_template: dropped the per-partner-language EN/FR branch —
CoC email attachments always render the single bilingual action_report_coc_en.
- fp_hide_default_reports: dropped the FR sequence record.
- Refreshed the report_coc.xml design note.
- Bump reports 19.0.11.32.0, notifications 19.0.7.1.0.
Deployed on entech (-u removed the orphan FR action + template). Verified the
cert Print menu now lists only "Certificate of Conformance" and it renders
clean. The dead action_report_coc_fr ref in the uninstalled
fusion_plating_bridge_mrp is left as-is (module not loaded).
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Changing Settings -> Certificate Owner didn't move existing certs: the signer
was snapshotted from the company owner at cert-creation time, and the CoC
prefers that snapshot over the live owner.
- _fp_create_certificates no longer freezes the company owner into
certified_by_id; it snapshots ONLY a deliberate per-spec signer. Empty
certified_by_id then resolves the LIVE company owner in the CoC report.
- action_issue lazy-fill made robust: resolves the company via the SO /
env.company (fp.certificate has no company_id) so it fills the CURRENT
owner at issue and the "Certified By" gate still passes.
- Settings help text corrected: signature comes from the user's Plating
Signature (Preferences -> My Profile), not "HR Employee".
- Data fix on entech: cleared certified_by_id on 5 stale draft CoCs with no
per-spec signer so they follow the current owner.
Bump certificates 19.0.9.3.0, jobs 19.0.11.4.0. Verified: CoC-30058 resolves
signer = Garry Singh (has Plating Signature), renders clean.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
The cert form's x_fc_local_thickness_pdf field only stored the upload; only
the Issue Certs wizard parsed it. Add create/write hooks on the jobs-side
fp.certificate that, when a NON-PDF is written to that field, run the wizard's
parser: readings -> thickness_reading_ids, header metadata -> x_fc_thickness_*,
microscope image (RTF) -> x_fc_thickness_image_id, then relocate the source to
x_fc_local_thickness_evidence_id and clear the PDF field (mirrors the wizard's
non-PDF end state). Real PDFs pass through untouched for the page-2 merge.
Re-entry guarded via the fp_skip_thickness_parse context flag. Bump jobs
19.0.11.3.0.
Deployed + verified on entech: CoC-30065 (.doc) back-filled to 3 readings +
metadata (operator BK) + extracted microscope image, renders inline (242KB);
PDF cert CoC-30040-02 correctly left untouched.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
- Column titles now render inline "English / French" on one line (was
stacked), cutting header height. First column drops "Line Item": it is
now Part Number / No. de pièce, Description / Description, Serial Number /
Numéro de série with a tight line-height.
- First-column DATA shows three lines — part number, part name, serial
number — via new fp.certificate._fp_resolve_part_identity() (part name
from the job's part catalog, serials from the matching SO line; blanks
fall back to "-"). Bump certificates 19.0.9.2.0, reports 19.0.11.31.0.
Deployed + verified on entech (CoC-30059: ('9876699373',
'VALVE BODY - COMPLETE - ASSY', ''), 243KB render).
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Make every CoC classic-body column title bilingual — English (bold) over
the French translation (italic grey), matching Steelhead and the SO
report's stacked-header convention. Cert-info headers (Date of
Certification / Generated By / Work Order #) and line-item headers
(Process / Customer PO / Shipped / NC Qty / Customer Job No.) now show
both languages. First column carries Part Number / Line Item,
Description, and Serial Number, each translated. Bump reports 19.0.11.29.0.
Deployed + render-verified on entech (CoC-30065, 224KB).
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
My prior change removed the .cert-statement-box border but the signature +
statement table was never bordered, leaving the whole section borderless.
Add class="bordered" so the two main columns (Certified By | Certification
Statement) get the outer box + divider like the other report tables; the
statement text keeps no separate inner box. Bump reports 19.0.11.28.2.
Deployed on entech (fusion_plating_reports upgrade clean).
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
The .cert-statement-box border was redundant next to the bordered tables;
render the statement as plain text (padding 0). Bump reports 19.0.11.28.1.
Deployed on entech (fusion_plating_reports upgrade clean).
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
- Customer block: remove the (usually empty) customer-logo third column;
Address | Contact now split 50/50.
- Remove the heavy header bottom border (the Sale Order header has none) and
the hr.heavy rule between the customer block and the cert info table.
- Drop now-dead CSS (.fp-coc h1, hr.heavy, .customer-logo). Bump reports to
19.0.11.28.0.
Deployed + render-verified on entech (CoC-30065, 222KB PDF, no QWeb errors).
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
- Drop the hard spec_reference gate on fp.certificate.action_issue. The
customer-facing description (_fp_resolve_customer_facing_description,
walks job -> SO line, reuses fp_customer_description) now drives the CoC
Process column; spec_reference prints only when an estimator fills it.
- CoC EN/FR reports swap web.external_layout for fp_external_layout_clean +
paperformat_fp_a4_portrait. New shared coc_header (company logo + address
left, Nadcap logo centre, title + Code128 barcode right) mirrors the Sale
Order header. Removed the 3-logo Nadcap/AS9100/CGP accreditation strip and
the body H1s; padding-top 0 on both body wrappers.
- Un-gate the Issue Certs wizard thickness upload (was invisible unless the
customer was thickness-flagged) so a Fischerscope report can be attached to
ANY cert; merge (page 2) + inline readings already render unconditionally.
- Update issue-gate tests, bump versions (certificates 19.0.9.1.0,
reports 19.0.11.27.0, jobs 19.0.11.2.0), record CLAUDE.md rule 14c.
Deployed + render-verified on entech (CoC-30065, 223KB PDF, no QWeb errors).
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
The NexaCloud->Odoo ledger now verifies every new invoice against its
SOURCE billing system before posting, instead of trusting NexaCloud's
unreliable created_at/status/paid_at:
- _fc_verify routes by stripe_invoice_id prefix (in_ -> Stripe REST,
lago: -> Lago REST) and returns source-truth
{invoice_date, void, draft, paid, paid_at, amount_paid}, or None when it
can't be determined/reached (left for the next run).
- _ingest_invoices(post=True, verified=...) uses the source invoice date
(and accounting date), and reconciles a payment ONLY when the source
confirms paid.
- _cron_sync_verified posts only finalized invoices; skips void + draft,
logs unverified for retry. Replaces the old _cron_ingest_recent.
Cron cron_fc_invoice_ledger is enabled daily on nexamain. First live run:
23 already-posted, 1 void + 2 Stripe drafts + 5 zero-amount all skipped,
0 new posted, ledger intact at $3,403.46.
Tests: routing/guards (no network), verified date+reconcile, and the cron's
void/draft/unverified filtering (sources patched). FCB_EXIT=0 on odoo-trial.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
_post_and_reconcile_paid: for invoices NexaCloud marks paid, set the ledger
entry's invoice_date AND accounting date to the original NexaCloud date,
post, then reconcile the Stripe payment dated to the actual paid_at. Unpaid
invoices stay draft. Per-invoice isolated. 76 tests green on odoo-trial.
One operator (e.g. "Gurpreet Singh") manages several distinct customer
businesses; naming partners from full_name mislabeled Mobility Specialties
Inc and Apex Vita Corporation as "Gurpreet Singh". Read the company field,
name the partner by company (mark is_company), and rewrite existing partners
so prior full_name-based names are corrected on re-ingest. 75 tests green.
Surfaced by the nexamain dry-run against real data:
- reader: cast invoice_items.invoice_id::text (uuid = text[] mismatch).
- readers: set_client_encoding('UTF8') — invoice descriptions contain "×".
- ingest: add a balancing line when invoice.subtotal != sum(items). 9 paid
base-plan invoices store the charge in subtotal with NO invoice_items, so
itemized ingestion under-recorded revenue by ~$1,143 (37%); the reconciling
line makes the Odoo invoice total match what Stripe billed.
74 tests green on odoo-trial.
Odoo becomes the accounting SoR by ingesting NexaCloud's real Stripe
invoices (read-only via the existing DSN) into native account.move
customer invoices: per-service-family income accounts, tax derived to
match the source invoice.tax, Stripe payments reconciled via
account.payment.register (invoice shows paid), idempotent on
x_fc_nexacloud_invoice_id, draft-first with bulk-post + a daily cron
(inactive). Plus a prune helper for the now-obsolete metered shadow data.
73 tests green on odoo-trial. Account codes use dots (Odoo 19 rejects '-').
Pivot from recompute-metered-billing to INGEST NexaCloud's real Stripe
invoices into Odoo account.move (posted + payment-reconciled + HST), driven
by the dual-run finding that 94% of NexaCloud revenue is Stripe service
invoices + add-ons + proration outside the per-deployment/CPU model. Full
accounting SoR, all history + ongoing, revenue split by service family,
draft-first rollout. Build/test on trial; reuses the read-only DSN + partner
mapping. Supersedes the metered direction for NexaCloud (engine kept inert).
Previous engagement notice used <blockquote> to style the findings
quote. Odoo's mail.thread renderer auto-tags every <blockquote> with
data-o-mail-quote-node="1" and the chatter UI then HIDES the content
behind a "..." widget — exactly the wrong UX since the findings are
the load-bearing content, not throwaway quoted text. Swapped both
quote blocks for styled <div>s with the same visual treatment (left
border, light background, padding) so they render fully inline with
no toggle.
Also expanded the notice to mirror more of what the owner sees in the
engagement email: now includes BOTH "Our reply" (the findings) and
"Summary sent to the owner" (the AI summary). The employee can see
the full context being used for the decision, not just the engineer's
reply. Skipped the Original Request section because the employee
wrote it themselves — would just clutter the thread.
white-space:pre-wrap preserves multi-line findings/summaries that the
engineer typed with line breaks. The two sections are visually
distinct: findings in light blue (matching the email's "Our Reply"
treatment), summary in light grey (matching "Summary for the
Decision" in the email).
Verified live on ticket #54: new message body has no <blockquote>,
no data-o-mail-quote attribute, and contains both section headers
with their content rendered inline.
Bumps fusion_helpdesk_central to 19.0.2.4.2.
Sending an engagement triggered template.send_mail(), which logged the
outbound email to the chatter as a `notification` message with the
internal `Note` subtype. That's correct for nexa-side bookkeeping (we
don't want the raw email body propagating to the customer), but it
meant nothing public was posted — so the entech-side My Tickets inbox
showed no activity. The employee couldn't tell their request had been
escalated for approval.
_fc_reset_engagement now posts a follow-up public message via
message_post (subtype mail.mt_comment, message_type='comment') with:
⏳ Awaiting owner approval from <owner_name>.
Their decision will appear here when they reply.
Our reply:
> <findings text>
This survives the entech _public_messages filter (comment +
non-internal subtype) and propagates to the employee's My Tickets
thread, giving them context AND the engineer's reply without exposing
the raw outbound email or the owner's email address.
Smoke-tested live on ticket #54: re-engaged with the same owner, the
new mail.message (id=348213) is subtype=Discussions / internal=False /
message_type=comment, and contains both the awaiting-approval notice
and the findings text. _public_messages would surface it.
Bumps fusion_helpdesk_central to 19.0.2.4.1.
The owner only saw the AI summary, which was a paraphrase of the user
report — they couldn't see the actual request OR what we said back.
Restructure the engagement email into three sections so the owner can
read the conversation and not just the AI's take:
1. Original Request (from the reporter) — ticket.description, no
longer buried in a <details> collapsible at the bottom
2. Our Reply — the wizard's "Your Findings" text, now persisted on
the ticket so the email template can render it directly. This is
the engineer's analysis / response to the request.
3. Summary for the Decision — the AI-generated brief
Approve / Reject buttons stay below all three. Bulk email mirrors the
same per-card structure.
New ticket field x_fc_engagement_findings (Text, copy=False) stores
the findings at send-time so they survive as audit history. Wizard's
_action_send_single / _action_send_bulk pass findings into
_fc_reset_engagement; bulk uses per-line findings + per-line summary.
Mail templates are in <data noupdate="1"> so a plain -u doesn't
re-import them. Pre-migration in migrations/19.0.2.4.0/pre-migration.py
deletes the existing template records + ir_model_data so the upgrade's
data load re-creates them with the new body_html. Pre- (not post-)
because data load happens between the two phases.
Smoke-tested live on nexa: rendered template HTML contains all three
section headers at the expected positions with their expected content
markers (ORIGINAL FROM RIYA in Original Request, REPLY-FROM-GURPREET
in Our Reply, the summary text in Summary for the Decision).
Bumps fusion_helpdesk_central to 19.0.2.4.0.
Previous fix (return True from action_generate_summary) prevented the
self-id crash but introduced a worse regression: Odoo's web client
auto-closes target='new' modals on any non-action return — the
"wizard done" convention. So Generate Summary updated the field and
then immediately killed the dialog, leaving the user with no chance
to click Send Engagement.
The only Odoo-19 idiom that reliably keeps a wizard dialog open
across a button click is to return an act_window dict that re-opens
the same wizard record (res_id=self.id). That was the original
approach — it crashed because of the active_id self-id collision in
default_get. With the active_model='helpdesk.ticket' guard now in
place (from 0104e877), the re-open is safe.
Belt-and-suspenders: the re-open action passes context={} explicitly,
so even if a future change to default_get drops the active_model
guard, there's no parent-form active_id leaking in to confuse the
ticket lookup. The wizard record is loaded by res_id directly; Odoo
19 doesn't call default_get for record loads, only for new-record
creation.
Centralised the re-open logic in _reopen_action so single + bulk
modes share the same code path.
Bumps fusion_helpdesk_central to 19.0.2.3.4.
- CRITICAL: reconciliation upsert keyed on (service, partner, period) collided
when one customer has two deployments (two subs) in a period — the second
overwrote the first. Add external_subscription_id to the model + a
UNIQUE(service_id, external_subscription_id, period) constraint, and key the
upsert per subscription. New test proves two subs for one partner keep two rows.
- raise a clear error if the nexacloud service is missing (was a confusing
per-row failure).
- _fc_resolve_subscription: the integer fallback no longer reaches a different
service's tagged subscription (latent multi-service IDOR); live untagged subs
stay resolvable and the partner-link authz is unchanged.
Full suite green on odoo-trial.
_api_record_usage now resolves the target subscription via the source
app's own id (x_fc_nexacloud_subscription_id, scoped to the service)
before falling back to a direct Odoo sale.order id. This is what lets
NexaCloud push usage against the shadow subscriptions the importer
created from NexaCloud UUIDs — closing the flip-day mapping gap the
review flagged. Authz unchanged (partner must be linked to the service).
fusion.billing.reconciliation gains the compute: _compute_reconciliation
(flat + charge overage vs external, status match/delta at a tolerance) and
_reconcile_rows (resolve shadow sub -> flat + charge, upsert one row per
service/partner/period, per-row isolated). The wizard gains a read-only
_read_reconciliation_rows (NexaCloud usage cpu_hours*3600 + invoice-item
subtotals per YYYY-MM) and a "Run Reconciliation" button. 2a amended to
stamp x_fc_nexacloud_plan_id on shadow subs so reconciliation can find the
charge. Read-only on NexaCloud; writes only reconciliation rows (shadow
guarantees intact). 8 new tests, full suite green on odoo-trial.
Repro: open the engagement wizard on a ticket, write findings, click
'Generate Summary from Findings'. Notification: "Ticket N no longer
exists" and the whole dialog closes — even though the ticket clearly
exists in the DB.
Root cause was two compounding bugs:
1. action_generate_summary returned an act_window dict with
res_id=self.id to "stay open after writing the summary field". The
web client honoured that by opening a NEW act_window — and the new
action's context inherited active_id=<wizard_id> (because that's
the res_id of the action being opened). Wizard ids are not ticket
ids, but our default_get didn't know the difference.
2. default_get read ctx.get('active_id') unconditionally, without
first checking ctx.get('active_model') == 'helpdesk.ticket'. So
when active_id pointed at the wizard's own id, default_get fed
that to _default_get_single, which raised "Ticket <wizard_id> no
longer exists" — and the user saw a confusing error about a
ticket that obviously DID exist (just not with that id).
Two fixes:
(a) action_generate_summary + action_generate_all_summaries now
return True. The form field write is visible to the client via
the call response; the wizard re-renders with the new
ai_summary populated. No spurious navigation, no context
pollution.
(b) default_get only consults active_id / active_ids when
active_model is helpdesk.ticket. Explicit
default_ticket_id[s] context keys still take precedence and
aren't gated by active_model (they're the caller's strong
signal).
Verified live: opening the wizard with active_id=99999 and NO
active_model no longer raises 'Ticket 99999 no longer exists' —
just creates the wizard cleanly. The normal flow (default_ticket_id
+ active_model='helpdesk.ticket') still works as before.
Bumps fusion_helpdesk_central to 19.0.2.3.3.
Previous fix (col=1 on the group) didn't work — Odoo still rendered
the group's string as a left-column label inside the form sheet's
flow, so the textarea got pushed into a narrow right column. The
summary field looked entirely missing because its content split
between the button row (on the right) and the textarea (collapsed
nowhere visible).
Right idiom (lifted from Odoo's own mail.compose.message wizard):
WIDE textareas live directly at the form level, not inside <group>.
Section titles use <separator string="…"> which renders as a
horizontal divider with the label above. The textarea then takes
the full sheet width naturally.
Same pattern applied to bulk mode for consistency.
Also moved Personal Note into the top compact group with Owner /
Owner Email since it's a one-line input that belongs with the
header info, not pretending to be a wide section.
Bumps fusion_helpdesk_central to 19.0.2.3.2.
Add action_test_connection — a read-only connectivity/schema check that
reports source row counts and imports nothing, the safe first step before
a dry-run. Wire a "Test Connection" button on the wizard. Document the
end-to-end run in the README: least-privilege read-only DB role SQL, the
fusion_billing.nexacloud_dsn system parameter (libpq DSN = NexaCloud's
URL minus +asyncpg), and the Test → dry-run → real-run flow. Refresh the
stale SCAFFOLD status. 53/53 green on odoo-trial.
User reported the rich text editor showing raw HTML tags as literal text
instead of rendering them as formatted prose. Root cause: Odoo's Editor
delegates content insertion to setElementContent() (web/core/utils/html.js),
which only takes the innerHTML branch when the content was flagged as safe
markup via owl's markup() helper. Plain strings fall through to the
textContent branch, which is what the user was seeing:
<p>Ask the client if the stairlift has power. Check:</p> <ul> <li>...
instead of the rendered paragraph + list.
The canonical html_field.js in @html_editor wraps its value with markup()
before passing it to the Wysiwyg config; I missed that detail.
FIX
- import markup from @odoo/owl
- in wysiwygConfig getter, wrap the saved content_html string with
markup() before assigning to config.content
- pass markup("") for empty content (avoids editor confusion with falsy)
- load-bearing comment to keep future refactors from re-introducing the bug
VERIFIED
- upgrade clean
- 7 stale asset bundles flushed, container restarted, login serves 200
- new bundle 014fee9 renders 10029808 bytes
- node --check PARSE_OK
- compiled bundle contains: content:rawHtml?markup(rawHtml):markup("")
which is exactly the markup-wrapped path the Editor wants
Bumped to 19.0.2.2.4.
Co-authored-by: Cursor <cursoragent@cursor.com>
The Findings and Summary fields rendered at half-width because their
enclosing <group> defaulted to col="2" — Odoo reserves a label column
even when the field has nolabel="1", so the textarea was squeezed
into the right half of the dialog while the left half sat empty.
Switch both groups to col="1" so the field uses the entire group
width. Also tag both fields with widget="text" explicitly (it was
inferred from the Text field type, but being explicit makes the
intent obvious to anyone reading the view) and migrate the button
row to a flex div so the helper text aligns with the button vertical
center.
Bumps fusion_helpdesk_central to 19.0.2.3.1.
The old flow fired OpenAI on wizard open with just ticket + chatter,
so the AI summary was just a paraphrase of what the user originally
reported — your engineering analysis (scope, limitations, recommended
approach) never made it to the owner. Restructure to a two-step flow:
1. Open wizard → empty findings + empty summary, NO OpenAI call
2. You write findings: scope / effort / approach / risk
3. Click 'Generate Summary from Findings' → OpenAI runs with
ticket + chatter + findings, where the prompt explicitly tells
the model to weight findings MORE THAN the original report
4. Review/edit, then Send
Bulk wizard mirrors the flow per line: each row gets its own
findings + summary, one 'Generate All Summaries' button fans out
parallel OpenAI calls using each line's own findings.
Updated SUMMARY_PROMPT to:
- Tell the model the support engineer's findings are authoritative
- Emit a bullet structure that leads with the recommendation, not
the user's restated ask
- Side with findings over the original report when they conflict
New tests cover:
- default_get does NOT fire OpenAI (regression guard for auto-AI)
- Findings text actually reaches the OpenAI prompt
- Send works with a manually-typed summary (no AI in the loop)
- Existing bulk + validation paths still pass with the new shape
Also folds in the deferred code-review #7: ThreadPoolExecutor now
explicitly cancels pending futures on timeout via
shutdown(wait=False, cancel_futures=True) so a slow OpenAI day can't
hold the wizard open for ceil(N/workers)*15s.
Bumps fusion_helpdesk_central to 19.0.2.3.0.
Smoke-tested live on nexa: opening the wizard makes zero OpenAI calls;
clicking Generate with findings='My findings: scope is XL, ~8h' makes
exactly one call and the findings text is verifiably in the prompt
body received by call_openai_chat.
REGRESSION FROM b22bb11b (Wysiwyg integration).
While inserting the new Wysiwyg methods (wysiwygConfig getter, onWysiwygLoad,
onToggleSource) between setup() and the existing onMounted / onWillUnmount
hook calls, I accidentally closed setup() early with the new
`this.wysiwygEditors = {};` assignment. That left the original
`onMounted(async () => {...});` and `onWillUnmount(...);` calls dangling
INSIDE the class body but OUTSIDE any method - which is invalid JS.
JavaScript's class-body parser sees the bare `onMounted(async () => ...)`
and tries to interpret it as a method declaration where `onMounted` is the
name and the parens are the parameter list. `async () => {...}` is not a
valid parameter, so the bundle fails with:
Uncaught SyntaxError: Unexpected token '('
web.assets_web.min.js:28807
That single parse failure tanks the entire backend asset bundle, leaving
users with a completely blank screen on /odoo (and any other backend
route). Frontend bundle was unaffected.
FIX
Move the onMounted / onWillUnmount calls back inside setup() where they
belong. Add a load-bearing comment explaining why they must stay there so
this regression cannot silently come back during a future refactor.
VERIFIED
- line 51: setup() opens
- lines 87, 93: onMounted, onWillUnmount calls INSIDE setup
- line 142: _initDrawflow as a normal class method (outside setup)
- upgrade clean
- bundle 10029245 bytes, exactly one onMounted( occurrence in
FlowchartDesigner class body
- node --check on the freshly-rendered web.assets_web.min.js -> PARSE_OK
Bumped to 19.0.2.2.3.
Co-authored-by: Cursor <cursoragent@cursor.com>
Resolves findings from the post-build review:
- C1: a partial import was indistinguishable from success. action_run_import
now logs failed rows at ERROR (survives nexa's log_level=warn) and the
wizard shows red/amber banners with failed/skipped counts.
- H3: an unrecognized billing_cycle silently fell back to monthly (wrong
plan AND price). Now raised per-row -> failed[], never silently mis-billed.
- M5: a NULL plan price silently became a $0 line. Prices now preserve
NULL-vs-0.0; a missing price for the subscription's cycle is failed[].
- H2: post-connect query/schema errors now become a clean UserError, not a
raw SQL traceback (matches the connection-error path).
- M4: per-row failures now record the exception type and log a traceback.
- MED#3: charge plan_id set explicitly False so re-runs re-assert the
shadow-safe NULL even if it was changed between runs.
- HIGH-edge: re-run only rewrites x_fc_* on existing subs; partner_id/plan_id/
line are set at creation only (never rewrite immutable fields).
- account_link: partner email match is now case-insensitive (=ilike) to avoid
duplicate partners against a differently-cased pre-existing partner.
Shadow-safety invariant unchanged and re-confirmed. 52/52 green on odoo-trial.
Adding the 'Fusion Helpdesk Central' block to General Settings so the
three ICP keys the engagement flow reads are configurable from a real
form instead of forcing admins to open Technical → System Parameters.
Three settings, all wired via config_parameter= so the existing read
paths (engagement_wizard, _fc_send_engagement_reminders) keep working
unchanged:
- fusion_helpdesk_central.openai_api_key (password widget — doesn't
render plaintext on the form)
- fusion_helpdesk_central.openai_model (default 'gpt-4o-mini')
- fusion_helpdesk_central.engagement_reminder_days (default 3, 0
disables the reminder cron entirely)
Bumps fusion_helpdesk_central to 19.0.2.2.0.
Find under Settings → Fusion Helpdesk Central. The block has two
sub-sections: "Owner Approval — AI Summary" (key + model) and
"Owner Approval — Reminder Cadence" (days).
fusion.billing.import.wizard backfills NexaCloud into Odoo: read-only
psycopg2 reader (_read_nexacloud_rows, DSN from ir.config_parameter)
split from pure-Odoo writes (_import_rows/_do_import) so the logic is
unit-tested headless. Maps users→partners+links (reusing
_resolve_or_create_partner, stashing stripe_customer_id), plans→a
cpu_seconds charge catalog (included_quota=cpu_seconds_quota,
unit_batch=3600, $0.0075/core-hour, plan_id NULL), and deployments→one
DRAFT shadow sale.order per deployment with the flat price set
explicitly. Shadow-safe by construction: draft + no payment token +
charge plan_id NULL (rating cron is a no-op). Idempotent re-runs;
per-row savepoints isolate bad rows; dry-run rolls back. 11 tests,
50/50 green on odoo-trial.
price_per_unit was a Monetary field, so a realistic sub-cent rate like
$0.0075/core-hour was rounded to $0.01 on write, corrupting the rate.
Make it Float(16,6). Also stop _compute_billable from rounding the
overage amount to 2 decimals mid-calc — that lost the half-cent on
sub-cent rates and would drift against the source app, which keeps
usage amounts at 4 decimals and only rounds at the invoice total.
Now rounds to 6 dp (float-noise only); cent-rounding defers to the
invoice line. Exposed while building the NexaCloud importer.
Adds a one-click 'loop the owner into the chatter' shortcut on the
ticket form — separate from the engagement approval flow, just keeps
the owner in the loop on ongoing communication.
What's new on helpdesk.ticket:
- x_fc_owner_display (computed Char): 'Kris Pathinather <kris@…>',
read live from fusion.helpdesk.client.key so a change to the owner
contact reflects immediately on every existing ticket.
- x_fc_owner_email_resolved (computed Char): email-only slice, drives
view visibility (the field + button only render when an owner is
configured).
- x_fc_owner_is_follower (computed Boolean): True when a partner with
the owner email is in message_partner_ids. Swaps the button for a
green 'Following' badge when the owner is already on the thread.
- action_add_owner_as_follower(): find-or-create the owner partner by
email and message_subscribe. Idempotent — second call is a no-op,
no duplicate partner. Raises UserError with a clear message if no
owner is configured.
View extension on the helpdesk ticket form: injects right after the
existing partner_id ('Customer') field in the customer side group,
so it reads as 'Customer | Owner Contact [Add as Follower]' — same
row, no layout shift when the state flips to 'Following'.
Tests cover the compute display in three states (configured,
no-client-label, no-owner-on-key), the action's three paths
(create-and-subscribe, reuse-existing-partner, idempotent-when-
already-following), and the UserError when nothing is configured.
Smoke-tested live on nexa: ticket with x_fc_client_label='ENTECH'
displays 'Kris Pathinather <kris@enplating.ca>'; first click adds
res.partner #723 to followers and flips owner_is_follower to True;
second click is a no-op.
Bumps fusion_helpdesk_central to 19.0.2.1.0.
Smaller UX simplification on the client side: the owner is already a
contact in entech's address book, so picking one is faster + safer than
re-typing their email and name (and avoids typos creeping into the
approval-email To: header).
What changed:
- Entech settings: drop fhd_owner_email + fhd_owner_name char fields;
add fhd_owner_partner_id Many2one to res.partner exposed in the
same "Owner Approval" block as a single partner selector. Quick-create
+ create-and-edit kept enabled so admins can spin up a new partner
inline if the owner isn't already in the system.
- controllers/main.py::_read_config: derives owner_email + owner_name
from the selected partner via the new _resolve_owner_contact helper.
Missing / dangling partner id → blank email + name → central simply
won't see the keys and the Engage button stays disabled (correct
"not configured" behaviour).
- Nexa side: ZERO changes. Still receives owner_email + owner_name
strings on the ticket payload, still upserts client_key.owner_email/
name. The partner abstraction stops at the entech boundary.
- migrations/19.0.2.1.0/post-migration.py auto-resolves the legacy
fusion_helpdesk.owner_email ICP value to an existing res.partner
(lowest-id match on lowercased email), writes the new
fusion_helpdesk.owner_partner_id key, and deletes the obsolete
owner_email + owner_name ICP rows so a future reader doesn't trip
over stale config.
Verified live on entech: kris@enplating.ca → res.partner #2308 ("Kris
Pathinather"), legacy keys purged, controller._resolve_owner_contact
returns the expected (email, name). The piggyback payload is unchanged
so existing client_key sync continues to work without a central
redeploy.
Bumps fusion_helpdesk to 19.0.2.1.0. fusion_helpdesk_central stays at
19.0.2.0.0 (no central-side changes required).
Replace the plain <textarea> in the flowchart designer's node-editor
right-panel with Odoo 19's native rich text editor so admins write
formatted prose / lists / bold / links / inline images without typing
HTML tags. The raw <textarea> stays available behind a toggle for the
power-user case (pasting markup from elsewhere, debugging).
CHANGES
manifest:
- depends += 'html_editor' (provides @html_editor/wysiwyg)
- bumped to 19.0.2.2.1
components/flowchart_designer/flowchart_designer.js:
- import { Wysiwyg } from '@html_editor/wysiwyg'
- import { MAIN_PLUGINS } from '@html_editor/plugin_sets'
- register Wysiwyg in static components
- state.sourceMode boolean (default false = rich text mode)
- wysiwygConfig getter builds the EditorConfig for the SELECTED node;
onChange reads editor.getContent() and writes back into the same
selectedMeta.content_html the rest of the designer already uses,
so the save path is unchanged
- onWysiwygLoad(editor) captures the editor instance per dfId so the
onChange callback can resolve the right one when nodes switch
- onToggleSource flushes the current editor's content before flipping
modes so unsaved keystrokes don't get lost
components/flowchart_designer/flowchart_designer.xml:
- replaced <textarea>...</textarea> with a conditional block:
sourceMode == false -> <Wysiwyg t-key="'wysiwyg-' + selectedNodeId"
config="wysiwygConfig"
onLoad="onWysiwygLoad.bind(this)"/>
sourceMode == true -> <textarea class="font-monospace" rows="10"/>
- t-key forces the editor to re-mount with the freshly-selected node's
content; otherwise switching nodes would keep showing the first
selected node's HTML
- new toolbar row above the editor has a "HTML Source" / "Rich Text"
toggle button (eye / code icons) so the user can flip at will
- hint text updated to reflect what each mode supports
components/flowchart_designer/flowchart_designer.scss:
- widened the right editor panel from 320px to 360px to give the
Wysiwyg toolbar room to breathe
- new .fr-wysiwyg-shell rule frames the embedded editor with the same
border + background as the other form-controls in the panel, with
a min-height of 180px and max-height 320px so it scrolls when the
content grows. Pins .o-we-toolbar inside the shell so it stays in
view as the user scrolls long content.
The save path, the runtime renderer, and the data model are unchanged -
content_html is still sanitised HTML stored on fusion.repair.flowchart.node.
Verified on local westin-v19:
- upgrade clean (no errors, no warnings)
- login serves 200 after restart
- 4 stale asset bundles flushed; Drawflow JS still served 46KB at
/fusion_repairs/static/src/lib/drawflow/drawflow.min.js
- Wysiwyg export confirmed at
/usr/lib/python3/dist-packages/odoo/addons/html_editor/static/src/wysiwyg.js:25
- MAIN_PLUGINS export confirmed at plugin_sets.js:103
Bumped to 19.0.2.2.1.
Co-authored-by: Cursor <cursoragent@cursor.com>
One-time, re-runnable, read-only importer that backfills NexaCloud
customers/plans/deployments into Odoo as a shadow copy for dual-run
reconciliation. Locks the brainstorming decisions: per-deployment
granularity, flat+overage billing, cpu_seconds metric, CPU-only v1,
Odoo-side psycopg2 reader, and shadow-safety by construction (draft
subs + no payment token + charges with NULL plan_id).
Findings from the post-feature code review on commit 396170b4. Addresses
the two CRITICAL + one HIGH + two MEDIUM issues; rest are deferred.
CRITICAL #1 — magic-link token race:
Two near-simultaneous POSTs on the same /engagement/<token>/approve
could both SELECT state='pending' under READ COMMITTED, both post
chatter, and let the last writer flip the outcome. Now the POST path
does an atomic UPDATE helpdesk_ticket SET token=NULL WHERE token=%s
AND state='pending' RETURNING id — the loser gets no row back and
renders the friendly invalid-link page. Verified live: 2 concurrent
POSTs → 1 wins, 1 loses, exactly 1 chatter row.
CRITICAL #2 — reminder cron without per-row savepoint:
Per CLAUDE.md rule #14, a DB failure mid-loop aborts the whole
transaction and silently kills the rest of the batch. Wrap each row's
send_mail+write in `with self.env.cr.savepoint()`. Also corrected the
success-count log (was len(stale), now actual sent count).
HIGH #3 — turnaround pivot summed instead of averaged:
fields.Float defaults to SUM aggregator; meaningless for per-ticket
decision delays. Added aggregator='avg' so the pivot reads "avg
turnaround per ticket" not "summed wait time".
HIGH #4 — added test_concurrent_claim_only_one_wins regression test
that fires two real HTTP POSTs against the same token and asserts
exactly one wins + exactly one approval chatter row exists.
MEDIUM #6 — cron nextcall pinned to 09:00 tomorrow so reminders land
in business hours regardless of when the module was last upgraded.
MEDIUM #10 — escalate failed owner-partner-create from WARNING to
ERROR (via _logger.exception) since silent attribution to the bot
account is a real audit-trail confusion.
Deferred (follow-up commits): #5, #7 (executor cleanup), #8, #9,
#11–#14 — none are bugs, all spec-drift or hardening.
Ships the design spec at docs/superpowers/specs/2026-05-27-owner-approval-flow-design.md.
What's new on central (fusion_helpdesk_central 19.0.1.2.0 -> 19.0.2.0.0):
- Engagement model: 8 new fields on helpdesk.ticket (state, snapshotted
owner email/name, single-use UUID4 token, sent/reminded/decided
timestamps, AI summary, stored-computed turnaround hours).
- Wizard: single + bulk modes on one fusion.helpdesk.engagement.wizard
TransientModel with a child wizard.line for per-ticket bulk summaries.
default_get pulls the OpenAI summary on open; AI fan-out for bulk is
parallel via ThreadPoolExecutor (max 5 workers, 30s overall cap).
- OpenAI client in utils.py — stdlib urllib, 15s per-call timeout, every
failure collapses to '' so the wizard's manual-summary fallback fires.
- Public portal: /fusion_helpdesk/engagement/<token>/<decision> GET +
POST, four branded standalone QWeb pages (confirm/done/invalid/error).
Token is single-use, cleared on confirm. Decision posts a public
comment attributed to the resolved owner partner; chatter propagates
to the employee's My Tickets thread per the "fully visible" UX choice.
- Mail templates (single + bulk) with magic-link buttons. Bulk template
renders one card per ticket, each with its own approve/reject URL.
- Reminder cron: daily, single-shot per engagement, configurable via
fusion_helpdesk_central.engagement_reminder_days ICP (default 3, 0
disables).
- Reporting dashboard: pivot/graph/list/kanban over helpdesk.ticket
filtered to engaged ones, with avg-turnaround measure. Menu lives
under Helpdesk > Reporting > Owner Engagements.
- Client_key extended with owner_email/owner_name fields; ticket.create
upserts them from the client-side piggyback (no new sync endpoint).
- 100% coverage on utils + integration tests on wizard, controllers,
re-engagement, cron, computed turnaround. OpenAI mocked in CI.
What's new on client (fusion_helpdesk 19.0.1.7.1 -> 19.0.2.0.0):
- Two new ICP settings: fusion_helpdesk.owner_email / .owner_name with
a new "Owner Approval" block in Settings > Fusion Helpdesk.
- controllers/main.py::submit piggybacks both keys on every ticket
payload so central keeps client_key.owner_email/name fresh
automatically.
Verified live end-to-end on entech -> nexa: payload upsert, wizard with
mocked AI, action_send, portal GET/POST/GET-again cycle, second click
hits the friendly invalid-token page. Token entropy = 122 bits (UUID4).
Two big workflow additions:
1. Visual drag-and-drop flowchart designer (Drawflow) + card-by-card runner
(with show-whole-tree toggle) so admins build per-(category, symptom)
decision trees with embedded photos/videos and CS walks callers through
them on the phone. Resolved-on-call closes the repair; escalated copies
the full transcript into internal_notes so the dispatched tech sees what
was already tried before they arrive at the client.
2. Vendor + draft-PO + factory-tracking on the part-order capture. Tech on
the phone with the factory picks the vendor from contacts, types the OEM
part #, cost, ETA date (calendar widget), factory ticket #, RA #, ticks
under_warranty, and the system auto-creates a draft purchase.order with
the right product (looked up or created from OEM) + activity for the
office on the ETA day + client email with ETA prominently shown and
cost intentionally omitted.
NEW MODELS
fusion.repair.symptom.class - lookup table (category + name + code).
Replaces the flat x_fc_issue_category Char on repair.order. Seeded with
7 stairlift symptoms + lighter coverage for hospital bed / porch lift /
lift chair. Equipment Class added to fusion.repair.product.category
(this carried over from the Bundle 10 plan).
fusion.repair.flowchart + .node + .edge - design-time graph.
- flowchart has name, category, symptom, version, published flag,
canvas_layout (Drawflow JSON), node_ids, edge_ids, computed start_node
- node has node_type (question / suggestion / info / outcome),
content_html, media_ids (M2M ir.attachment for photos + videos),
is_start, outcome_kind (resolved / escalate / order_part),
canvas_x/y for Drawflow round-trip
- edge has source, target, label, sequence - supports N-ary branching
(not just Yes/No)
- designer_load() and designer_save(payload) RPC API the OWL component
consumes; save is atomic-replace + bumps version + soft-validates
fusion.repair.flowchart.run + .step - runtime sessions.
- One run per repair, audited; runtime_start_or_resume() returns the
existing in-progress run or creates a fresh one for the matching chart
- runtime_choose(edge_id, cs_note) records a step + advances current_node
- runtime_complete(outcome) snapshots final node + calls _apply_outcome:
resolved -> auto-close via action_repair_start + action_repair_end,
set x_fc_resolved_on_call, post transcript to chatter
escalated -> prepend transcript to repair.internal_notes so the tech
sees it first when they open the form
order_part -> chatter note; tech opens visit-report wizard next
abandoned -> just save transcript
- Each step snapshots node_name + chosen_label at write time so the
transcript survives later chart edits without breaking.
REPAIR.ORDER EXTENSIONS
- x_fc_symptom_class_id (M2O) - new structured symptom field
- x_fc_resolved_on_call (Boolean, tracked) - true after a resolved outcome
- x_fc_flowchart_run_ids + x_fc_flowchart_run_count
- action_start_troubleshoot() - opens the runner client action, raises a
helpful UserError if no symptom set or no published chart exists
- action_view_flowchart_runs() smart button
- x_fc_issue_category renamed string to "(legacy)" - kept for back-compat
+ AI prompt context; new intakes set the M2O
DRAWFLOW DESIGNER (OWL)
static/src/lib/drawflow/drawflow.min.{js,css} - vendored Drawflow 0.0.59
(MIT). Loaded only in web.assets_backend, ~48KB total.
components/flowchart_designer/flowchart_designer.{js,xml,scss}:
- Client action "fusion_repair_flowchart_designer" with full drag-drop
canvas + zoom + pan
- 4 custom node templates color-banded by type (question blue,
suggestion green, info gray, outcome red/green/amber per outcome_kind)
- Right-panel editor for selected node: title, type, outcome kind,
content (HTML), media uploader (drag-drop or click), set-as-start
toggle, per-outgoing-edge label editor
- Save serializes Drawflow JSON to canvas_layout + atomic-replaces the
structured node/edge rows via the designer_save RPC
CARD RUNNER (OWL)
components/flowchart_runner/flowchart_runner.{js,xml,scss}:
- Client action "fusion_repair_flowchart_runner"
- DEFAULT MODE: card-by-card. One big card per node, embedded photos +
inline <video controls>, answer buttons sized for phone use, CS note
textarea (saved as cs_note on the step), running transcript at the
bottom
- TOGGLE: "Show Whole Tree" loads the same Drawflow lib in read-only
fixed mode, imports the canvas_layout JSON, highlights current node
yellow / visited green via .fr-current / .fr-visited classes
- Outcome buttons drive the right runtime_complete() call; success
notifications + auto-return to the parent repair form
- "Abandon & Escalate" header button at all times - transcript is saved
even on bail-out so the dispatched tech still benefits
PART ORDER + VENDOR PO
repair.part.order new fields:
vendor_partner_id (M2O res.partner, is_company domain), purchase_order_id
(auto-created draft PO), product_id (auto-resolved or created),
unit_cost (Monetary) + currency_id, internal_po_ref, factory_ticket_ref,
factory_ra_number, under_warranty.
action_create_draft_po() - resolves product.product by OEM (default_code)
or creates a new one in a "Spare Parts" product.category, creates a
purchase.order in draft state with one line (product + qty + price_unit
+ date_planned from expected_date or +7d), stamps Westin's internal PO
ref as partner_ref so the factory can find it on return. Office reviews
and confirms via the normal Odoo flow.
_schedule_eta_activity() - schedules a Repair: Assign Technician activity
on the parent repair.order due on expected_date, assigned to
repair.user_id, so the office is reminded to call the client and book
the return visit on the day parts arrive.
VISIT-REPORT WIZARD PARTLINE EXTENSIONS
Same new fields exposed inline on the partline list so the tech captures
everything on the phone with the factory in one form:
vendor_partner_id (vendors-only filter), unit_cost + currency,
expected_date (calendar widget) replacing expected_lead_days as the
preferred input, under_warranty, internal_po_ref, factory_ticket_ref,
factory_ra_number, create_draft_po (default True - auto-builds PO on
submit when vendor + cost are both set).
CLIENT EMAIL TIGHTENED
email_template_parts_ordered:
- Subject now includes ETA "Parts ordered for your stairlift - expected 2026-06-06"
- Hero ETA panel: large blue-bordered card with "Expected Arrival" label
and the date in 24px bold
- Cost INTENTIONALLY OMITTED - "Our office will call you to confirm a
return visit time. If you have any questions about pricing or
scheduling, please reach out to our office directly."
- "There is nothing for you to do right now." callout
UI
- repair.order form header: new "Start Troubleshooting" button (info
style, sitemap icon, visible when state in (draft, confirmed,
under_repair) AND symptom is set)
- repair.order form intake row: x_fc_symptom_class_id picker filtered to
the category, x_fc_resolved_on_call display when true
- repair.part.order form: header button "Create Draft Purchase Order"
+ new Vendor / Cost / Warranty group + System group with the PO link
- Intake wizard equipment line: symptom_class_id picker
- New menus:
Configuration > Symptom Classes
Configuration > Troubleshooting Flowcharts
Fusion Repairs > Troubleshooting Sessions (run history)
SECURITY
18 new ACL rows for the 6 new models, scoped Manager-full / User-read /
FieldTech-read. Flowchart runs and steps get write access for User so CS
can record steps; Manager owns flowchart + node + edge CRUD.
POST-MIGRATION (19.0.2.2.0)
Existing installs: walks all distinct (category, x_fc_issue_category) text
pairs on repair.order, creates a placeholder fusion.repair.symptom.class
per pair (or reuses an existing match by code/name), back-fills the new
x_fc_symptom_class_id M2O. Idempotent + safe to re-run.
DEPENDENCY
Added 'purchase' to depends (action_create_draft_po needs purchase.order).
VERIFIED END-TO-END on local westin-v19 (Margaret persona, 0 bugs):
STEP 0 seed: chart v1 8 nodes / 12 edges / published, 7 stairlift
symptoms, stairlift class=lift_elevating
STEP 1 CS creates RO-202605-60 with symptom Not Moving
STEP 2 Start Troubleshooting -> client action tag returned
STEP 3 walk run: Power on? Yes -> Seatbelt? Yes -> Swivel? Yes ->
outcome 'Still not moving - dispatch technician'
(outcome_kind=escalate)
STEP 4 runtime_complete('escalated') -> internal_notes prepended with
CS troubleshooting summary
STEP 5 visit-report parts_needed with vendor Handicare + cost $425 +
warranty + factory refs -> PART-00008 created + draft
PO 26690 auto-built with line "Handicare 1100 control
board" qty 1 @ $425, partner_ref WH-2026-1042
STEP 6 mark_ordered -> client email queued (NO cost mentioned, ETA
shown prominently) + office activity scheduled for
2026-06-06
STEP 7 fresh resume returns same run; resolved outcome auto-closes the
repair (state=done, x_fc_resolved_on_call=True)
Bumped to 19.0.2.2.0.
Co-authored-by: Cursor <cursoragent@cursor.com>
End-to-end spec for the owner-approval feature on fusion_helpdesk +
fusion_helpdesk_central. Captures data model, engagement flow (single +
bulk), magic-link approval portal, OpenAI summary, reminder cron,
reporting dashboard, edge cases, and test plan. Ready for the
writing-plans skill to turn into an implementation plan.
The OWL dialog used <t t-out="m.body"/> on message bodies, but t-out
escapes plain strings — it only renders raw when the value is a Markup
instance. Bodies arrive over JSON-RPC as plain strings (Markup is a
client-side type, doesn't cross the wire), so the customer was seeing
literal "<p>This has been fixed.</p>" in the thread instead of the
rendered HTML.
Wrap incoming bodies in `markup()` at the boundary (openTicket +
sendReply call sites) so the template renders them as the sanitised
HTML the central chatter already produced. Trust is fine — the body is
sanitised server-side by mail.thread before it ever leaves nexa.
Bumps fusion_helpdesk to 19.0.1.7.1.
Three coordinated changes on top of the section grouping:
1. **Mark as Critical** — a red chip on the New tab sets priority='3'
when submitted. The central post-create hook auto-applies a "Critical"
helpdesk.tag (shipped via fusion_helpdesk_central data XML, noupdate=1
so support can recolor without losing it on upgrade), giving support
a kanban-groupable signal that doesn't rely on remembering what
priority='3' means. Scoped to in-app-channel tickets only, so a
support agent manually setting Urgent on their own ticket isn't
silently tagged.
2. **KPI cards above the sections** — Total / Open / Closed / Critical
in a 4-up grid (auto-collapses to 2x2 under 540px). Each card uses
its own saturated gradient so it reads on both light and dark mode —
the dialog backdrop is irrelevant because the gradient brings its
own background. Counts are computed in JS from state.tickets so they
always match what's rendered below.
3. **Colored stage pills** — red Critical, green Solved, dark-yellow New,
orange Cancelled, blue for In Progress / Testing / On Hold. Critical
priority gets a *separate* red pill alongside the stage pill so you
keep stage info even on escalated tickets. Stage matching is
substring-based (lowercased) so a renamed "Resolved" or "Done" stage
on central still maps to the green pill.
Tests cover the new is_critical=True → priority='3' wiring and the
default omission so SLA / stage defaults keep working for normal
tickets. Bumps fusion_helpdesk to 19.0.1.7.0 and
fusion_helpdesk_central to 19.0.1.2.0. End-to-end smoke test verified
live: priority=3 + x_fc_client_label triggers the Critical tag.
The flat write_date-sorted list was hard to scan with 50+ tickets — solved
ones were intermixed with active ones, and there was no signal for
priority. Bucket each ticket server-side into 'critical' (open + priority
High/Urgent), 'solved' (stage marked fold=True on central) or 'open'
(everything else), and render three labelled sections in the dialog with
sticky headers, count badges, and per-group accent colours. Backend keeps
its write_date desc order so latest is always at top within each bucket.
Bucketing uses helpdesk.stage.fold (not the stage name) so renaming
"Solved" to "Done" on the central won't quietly mis-categorise rows.
Adds bucket_ticket() in utils.py with unit tests covering the
folded-wins-over-priority precedence and the missing-priority fallback.
Also surfaces a small Urgent (triangle) / High (arrow) icon on each row
so a critical ticket reads at a glance even after a user scrolls past
the section header.
Bumps fusion_helpdesk to 19.0.1.6.0.
The customer-followup ship left two papercuts that hid 51 historical
tickets from the entech owner:
1. group_reporter_admin had zero members on install — the new XML record
created the group but never granted it. Extend base.group_system's
implied_ids so every system administrator transparently inherits the
admin view of the embedded inbox on install / upgrade. (4, id) tuple
is additive — never replaces base's existing implications.
2. Tickets created before this feature shipped had NULL
x_fc_client_label and NULL partner_email, so the scope filter
excluded them all. The reporter identity was still recoverable from
the description HTML's diag block. Backfill recipe is captured in
CLAUDE.md so future deployments can apply the same one-shot UPDATE
without re-deriving the regex.
Bumps fusion_helpdesk to 19.0.1.5.0. Verified live on entech: all six
base.group_system members now return True for
has_group('fusion_helpdesk.group_reporter_admin').
Adds two Integer fields to res.partner:
- x_fc_default_lead_time_min_days
- x_fc_default_lead_time_max_days
Set once on the customer's Plating Defaults tab (Fulfilment group);
auto-copies onto every new Express Order via the existing
_onchange_partner_id hook. Operator can still override per-order
since the onchange only fills when the wizard field is still blank.
Field declaration lives in fusion_plating_configurator (alongside
the rest of the partner cascade reads). View edit lives in
fusion_plating_invoicing where the Plating Defaults tab already
hosts the other partner-level defaults (invoice strategy, deposit
%, delivery method, deadline-days). Invoicing depends on
configurator, so the fields are registered before the view loads.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Three related fixes on the Express Orders totals card:
1. Totals card now breaks out Subtotal / Tax / Tooling Charge /
Grand Total. Previously the "Subtotal" and "Grand Total" rows
both read from total_amount (same value rendered twice) and no
tax was shown at all. Customers on a fiscal position-mapped
tax rate (Ontario HST, etc.) had their taxes silently dropped
from the preview.
2. tooling_charge now feeds the Grand Total. The total_amount
compute previously summed line subtotals only. Added a real
SO line for the tooling charge in action_create_order so the
eventual sale.order.amount_total matches the preview AND the
invoice carries a "Tooling Charge" line item.
3. tax_ids is now visible as an optional column on the lines
list. Operator can see + override the auto-applied tax per
line. Default still comes from FP-SERVICE product mapped
through partner.property_account_position_id (fiscal position).
New compute fields on fp.direct.order.wizard:
- total_subtotal (sum of line.qty * line.unit_price, pre-tax)
- total_tax (sum of line + tooling taxes via compute_all)
- total_amount (subtotal + tax + tooling — was just subtotal)
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Durable: nexa/entech upgrade commands, central service-account Contact
Creation prerequisite, backup-outside-addons-path gotcha, smoke-tests-must-
call-the-controller lesson. Plus current deploy status + the one remaining
step (browser confirmation of My Tickets / New on entech).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Squash-merge of feat/helpdesk-customer-followup. The billing and
fusion_login_audit work from that branch is already on main (landed
separately); this lands only the helpdesk feature.
- Identity keystone: submit() forwards partner_email/partner_name/
x_fc_client_label so the central Helpdesk find-or-creates the customer
partner and subscribes them as a follower (enables reply emails + magic link).
- Embedded in-app 'My Tickets' inbox: server-side scoped read/reply RPC
endpoints, per-user seen tracking (fusion.helpdesk.ticket.seen), systray
unread badge. Defense-in-depth scope domain + _norm_email normalisation
(wildcard emails cannot widen scope).
- fusion_helpdesk_central: x_fc_client_label field + list/search views +
branded acknowledgement email template.
- Deployed and smoke-tested live: nexa central 19.0.1.1.0, entech client
19.0.1.4.1 (requires Contact Creation on the central service account).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
When the login string resolves to an existing user and the password is
wrong, BOTH overrides used to write a failure row:
- _check_credentials wrapper: result=failure, reason=bad_password
- _login wrapper (catching the propagating AccessDenied): result=
failure, reason=unknown_user
Discovered in production smoke on westin-v19 after the deploy: a
single failed login for info@gsafinancialconsulting.com produced two
audit rows (one bad_password, one unknown_user). The unknown_user
label was wrong — the user IS in the system.
Fix: _login now checks whether the login string resolves to any user
BEFORE writing the unknown_user row. If yes, _check_credentials
already logged the attempt and _login skips. If no, the user lookup
in super() failed and _login is the only chance to log.
Regression test test_login_known_user_bad_password_single_row asserts
exactly one row per attempt and that the row carries bad_password
(not unknown_user) when the user exists.
30 tests green locally; production smoke on westin-v19 confirms:
one row per failed login, bad_password, IP 172.18.0.1 captured.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Capture in the plan the Odoo 19 gotchas discovered during execution
that the original plan template missed:
- Test command requires --http-port=0 --gevent-port=0 (running
container holds 8069).
- Declarative models.Constraint / models.Index (T2).
- res.users.groups_id renamed to group_ids (T3, T6).
- ir.rule groups is additive not restrictive (T3).
- mail.template inline-template ctx IS env.context (T11).
- ir.cron has no numbercall field in 19 (T12).
- registry.cursor() in tests is TestCursor; cr.commit() raises;
use savepoints (T13).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Asserts the smart-button and Login Activity tab fields are stripped
from res.users get_view() for non-admin users, and present for
Settings admins. Locks down the contract behind the
groups="base.group_system" XML attributes on the form-inheritance
view (the inherited view record cannot carry groups itself per
CLAUDE.md rule #11; the gate must live on the inner nodes).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
5-min cron processes up to 100 pending rows per pass: private IPs
short-circuit to state=private_ip; same-IP cache (30 days) avoids
duplicate ip-api.com calls; reverse DNS via socket with 1.5s timeout;
HTTP lookup respects ip-api''s X-Rl rate-limit header. Tests cover
private-IP shortcut, cache hit (no HTTP), and internal-state skip --
no network calls needed.
Per-row isolation uses cr.savepoint() instead of cr.commit() because
Odoo 19 TestCursor raises AssertionError on commit/rollback. Recorded
the gotcha as CLAUDE.md rule #14.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Adds _fc_retention_gc() that deletes rows older than the configured
horizon (default 365 days; 0 = keep forever). Registered as a daily
ir.cron. Tests verify both the delete path and the "keep forever"
short-circuit.
Also documents the Odoo 19 gotcha that ir.cron dropped the numbercall
field (the legacy "-1 = run forever" pattern now raises ValueError at
install time; just omit the field).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Mail template + helpers (_fc_alert_*, _fc_recent_failure_count,
_fc_send_failure_alert) wired into _check_credentials so that crossing
the consecutive-failure threshold within the window queues exactly one
mail.mail per attempted login per 60-minute cooldown. Master switch
x_fc_login_audit_alert_enabled honoured. Recipients are members of
base.group_system with a non-empty email and share=False; the
__system__ superuser is excluded by Odoo''s default user filter.
Tests (3 new, 22 total green):
test_failure_burst_queues_one_email
test_cooldown_suppresses_second_alert
test_alert_disabled_master_switch
setUp ensures base.user_admin has an email (fusion-dev''s admin user
ships without one; the only user with an email is __system__, which
is filtered out of standard res.users searches).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Four x_fc_* fields on res.config.settings backed by ir.config_parameter:
retention_days (default 365, 0 = forever), alert_threshold (5),
alert_window_min (15), alert_enabled (True). New "Login Audit" block
on the General Settings page (gated by base.group_system on the block,
NOT on the inherited view record per CLAUDE.md rule #11).
CLAUDE.md gotchas added during this task:
#5 Boolean config_parameter fields don't round-trip "False" as a
string — IrConfigParameter.set_param deletes the row on falsy.
Test with assertFalse, never assertEqual(..., "False").
#6 ir.ui.view uses group_ids (Odoo 19 rename mirrored from res.users).
Setting groups_id on an ir.ui.view record raises ValueError at
install. (The XML attribute groups="..." on inner nodes is
unrelated and still works.)
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
List, form, and search views for fusion.login.audit, plus a "Login
Events" full-history action and a "Failed Logins (24h)" pre-filtered
action. Both surface under Settings -> Technical -> Login Audit
(menu items gated by base.group_system). Views are no-create / no-edit
/ no-delete to enforce append-only at the UI layer too.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Adds four x_fc_* fields on res.users: login_audit_ids (One2many),
login_audit_count (compute), last_successful_login (compute, stored),
last_login_ip (compute, stored). action_fc_view_login_audit returns
a window action scoped to the current user. View inheritance adds a
smart button to the button box and a "Login Activity" page to the
notebook, both gated by base.group_system on the inner XML nodes
(NOT on the view record — Odoo 19 forbids that; see CLAUDE.md rule #11).
Tests (2 new, 18 total green):
test_computed_last_successful_login — uses registry cursor to commit
the audit row so the stored compute picks it up across the
TransactionCase boundary.
test_action_view_login_audit_returns_window_action — smart-button
action shape + domain scoping.
CLAUDE.md rule #11 added: inherited ir.ui.view records cannot have
groups/group_ids on the record; the gate must be on the inner XML nodes.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Overrides res.users._login. When the login string does not resolve to
any user, super() raises AccessDenied; we record a row with user_id=NULL
and failure_reason="unknown_user", then re-raise. Closes the gap where
typo'd or scanned logins would otherwise vanish from the audit trail.
The existing _fc_record_login_event helper writes through an independent
registry.cursor(), so the audit row survives the rollback that follows
the re-raised AccessDenied.
Note: in Odoo 19 _login is a plain instance method (not the classmethod
it was in earlier versions) and takes (credential, user_agent_env). The
original plan was written for the classmethod signature; corrected here
and recorded in CLAUDE.md rule #10 so future-Claude does not waste time
re-discovering it.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Wraps res.users._check_credentials. On AccessDenied, records a row with
result=failure and failure_reason='bad_password' (or '2fa_failed' when
credential['type'] == 'totp'), then re-raises. Regression test asserts
the attempted password value never lands in any audit field.
The audit row is written through registry.cursor() (independent cursor) so
it survives the rollback that follows AccessDenied — in production
odoo/service/model.py::retrying resets the transaction and http.py closes
the cursor without committing, in tests assertRaises opens its own
savepoint. Either way an inline write would vanish. Tests
enter registry_test_mode and use manual try/except to keep the audit row
visible across the savepoint hierarchy.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Overrides res.users._update_last_login to create a fusion.login.audit
row with result=success after the parent runs. The write goes through
sudo() + mail_create_nolog=True. Any exception in the audit path is
caught and logged but never propagates — a broken audit table must
never block a real user from logging in.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Single helper builds vals for fusion.login.audit rows from the live
HTTP request, or falls back to ip=''internal'' + geo_lookup_state=''internal''
when there is no request. Parses UA into browser/os/device_type via the
bundled user_agents library. Never reads credential[''password'']. Tests
cover: no-request fallback, UA parsing on a Chrome/Windows UA, and the
regression that no password value leaks into the vals dict.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Record rule grants admins an unrestricted domain on the audit log;
ACL forbids write/create/unlink for every group (audit is append-only;
sudo() inside auth hooks is the only write path). Defence-in-depth
layering: ACL is the actual gate, the rule documents and locks down
admin access path.
Tests (5, all green) cover:
test_admin_can_read_through_acl_and_rule — positive path through both.
test_acl_blocks_read_for_regular_user — base.group_user denied by ACL.
test_acl_blocks_read_for_portal_user — base.group_portal share user
denied (sensitive data leakage
surface closed at ACL layer).
test_acl_blocks_write_for_admin — append-only at the write boundary.
test_acl_blocks_unlink_for_admin — append-only at the unlink boundary.
Drop the redundant `from . import tests` from the root __init__.py —
Odoo's test loader imports `odoo.addons.<mod>.tests` directly; the
extra import was dead weight (and inconsistent with the repo pattern).
CLAUDE.md gotchas added during this task:
#6 res.users.groups_id -> group_ids rename (test setUp pitfall).
#6 ir.rule `groups` is additive, not restrictive — group-scoped
rules only apply to users in that group, they do not restrict
non-members. Default to letting the ACL gate; use rules for
row-level filters ACLs cannot express.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
- All 16 columns per spec (user, attempted_login, result, failure_reason,
event_time, ip/geo fields, user_agent triple, device_type, database).
- Check constraint binds failure_reason presence to result value.
- Three composite indexes (user+time, login+time, geo_state+time) supporting
the per-user, failure-burst, and geo cron queries.
- Minimal admin-read ACL added so subsequent tests can verify writes.
- 3 TransactionCase tests passing: model create, failure_reason nullable on
success, geo_lookup_state='internal' accepted.
Odoo 19 deprecation note: this implementation uses the declarative
models.Constraint and models.Index attributes (Odoo 19 silently drops the
legacy `_sql_constraints = [...]` list and `init()`/raw-SQL pattern with
only a warning). Captured in CLAUDE.md rule #9.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Empty installable module with manifest, package inits, and icon.
Subsequent tasks add the audit model, hooks, views, and tests.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Durable login audit for Odoo 19 (westin-v19). Captures successful and
failed authentications via _update_last_login / _check_credentials /
_login overrides, surfaces history on res.users as a smart button +
"Login Activity" tab (admins-only), async geo-enriches IPs via ip-api.com
through network_logger, 365-day retention with daily GC cron, and
emails Settings admins on N consecutive failures for the same login
within a configurable window.
Motivation: a spot audit of GSA Accounting (uid 63) showed Odoo's
res_users_log keeps only one row per user (rest is GC'd), /var/log/odoo
is empty (warn-level stdout logging), and the container json log
rotates within days — leaving no durable login trail.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
docs(billing): session handoff — core on main, sub-project #2 (NexaCloud) next
Captures resume state for the centralized-billing initiative: core engine done
and on main, the 4-chunk decomposition of sub-project #2 (NexaCloud adapter +
dual-run reconciliation), the pending "where to start" decision, open questions,
and the test/branch workflow.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
@
Add _match_api_key() class method to fusion.billing.service, with a
TDD test suite (TestServiceApiKey) covering key generation, hash storage,
positive match, and rejection of bad/inactive keys. Also fix
fcb_test_on_trial.sh to use --http-port 8070, as Odoo 19 forces
http_spawn() even under --no-http when --test-enable is set.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Local dev Odoo is Community (can't install the module). Add a guest-exec runner
that syncs the module to the odoo-trial Enterprise sandbox (VM 316, db trial) and
runs --test-enable there; pass = FCB_EXIT=0. Scaffold verified installing on
Odoo 19.0 Enterprise (7 fusion_billing_* tables created).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Centralize billing for all NexaSystems services (NexaCloud, NexaDesk,
NexaMaps, custom apps, memberships) on the Odoo 19 Enterprise instance,
replacing Lago. The module adds only the metering + integration layer;
native sale_subscription / account_accountant / payment_stripe do all the
financial work (invoicing, HST, dunning, portal, credit notes, Stripe).
Includes:
- Design spec (docs/superpowers/specs/2026-05-27-nexa-billing-centralized-design.md):
6 locked decisions, architecture, data model, usage engine, Lago-shaped
API, webhook control loop, NexaCloud pilot, phased dual-run migration.
- Module scaffold: 7 fusion.billing.* models (service, account.link, metric,
charge, usage, webhook, reconciliation), bearer-auth API controller shell,
security ACLs, README. Compiles on Odoo 19.0; engine/API bodies are stubs
pending the implementation plan.
- CLAUDE.md rule #15: no sale.subscription model in Odoo 19 — a subscription
is a sale.order(is_subscription) + sale.subscription.plan (verified live).
Task 0 verified: a single Stripe account is shared across NexaCloud and all
Lago providers, so no Stripe account/card migration is required.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
fp.step.template rows already held 'fa-bathtub' (1), 'fa-flag' (2),
and 'fa-undo' (2) — all plating-relevant and presumably valid in an
earlier version of the Selection list. When step_insert snapshot-
copied these into a fresh fusion.plating.process.node via
_copy_snapshot_fields, the ORM rejected them with
ValueError: Wrong value for fusion.plating.process.node.icon
because they weren't in the curated 39-icon list anymore.
Adding 'fa-bathtub' (bathtub / tank / soak), 'fa-flag' (flag /
milestone / gate), and 'fa-undo' (undo / rework / rerun) to the
process.node Selection. Aligns the two lists (template uses
_get_icon_selection -> node._fields['icon'].selection at runtime).
No data migration needed — existing template rows immediately
re-validate against the wider Selection.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
FpExpressActionBtns.onOpen called action_open_part which returned an
ir.actions.act_window dict without a 'views' key. Odoo 19's
_preprocessAction in the web client tries to .map over action.views
and throws TypeError: Cannot read properties of undefined (reading 'map').
Fix: include 'views': [[False, 'form']] alongside view_mode='form' on
both copies of action_open_part (wizard line + sale.order.line).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Three view edits to surface the new cert toggles + workflow nudges:
1. res.partner — Plating Documents tab gains a "Aerospace / Defence"
separator + group with the three new toggles (Nadcap / MTR /
Customer-Specific). All boolean_toggle widget, default OFF.
2. fp.process.node — Recipe form gains a "Certificate Output" group
visible only when node_type == 'recipe'. Five requires_* toggles
+ a blue info banner explaining the suppress-only precedence.
3. fp.certificate — Certificate PDF tab gains a yellow alert banner
when certificate_type is one of the three orphan types AND no
attachment is set. Tells the operator "this type expects a PDF
you upload from disk".
Sub: docs/superpowers/specs/2026-05-27-recipe-cert-toggles-design.md
Task: T6.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Block fp.certificate.action_issue on Nadcap / Mill Test / Customer-
Specific certs when attachment_id is empty. These three cert types
are manual-attach only — operator uploads the supplier doc /
regulator-issued cert / filled customer template PDF before the
cert can be issued. Prevents shipping the customer an empty PDF.
_fp_render_and_attach_pdf gets an early-return guard so an orphan-
type cert never tries to render a CoC QWeb template.
Sub: docs/superpowers/specs/2026-05-27-recipe-cert-toggles-design.md
Task: T5. Makes test_orphan_cert_issue_blocks_without_attachment pass.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Rewrites fp.job._resolve_required_cert_types as a documented three-step
pipeline:
Step 1 — partner + part flags (extended to read 3 new orphan-type
partner toggles: x_fc_send_nadcap_cert / x_fc_send_mill_test
/ x_fc_send_customer_specific)
Step 2 — recipe-level requires_* Booleans STRIP cert types from
the wanted set (suppress-only — never adds)
Step 3 — CoC + thickness bundling preserved (thickness collapses
into CoC PDF as page 2)
Field-existence guards on partner/recipe attribute reads keep the
resolver robust if the certificates / plating module schemas drift.
Recipe is suppress-only per Q1 locked decision: customer/part is the
ceiling, recipe can only remove. Test 3 (test_recipe_cannot_add_certs_
customer_didnt_want) is the explicit regression guard.
Sub: docs/superpowers/specs/2026-05-27-recipe-cert-toggles-design.md
Task: T4. Makes the 5 resolver tests from T3 pass.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Six failing tests in test_recipe_cert_suppression.py covering the
full design surface:
1. test_recipe_suppresses_thickness
2. test_recipe_suppresses_nadcap_for_commodity_part
3. test_recipe_cannot_add_certs_customer_didnt_want (suppress-only
regression guard — recipe can never add types customer didn't ask for)
4. test_part_override_coc_recipe_suppresses
5. test_all_orphan_types_propagate (4-element output + bundling)
6. test_orphan_cert_issue_blocks_without_attachment
These will all fail until T4 (resolver) and T5 (orphan-attach gate)
land. RED phase of TDD locked in via commit ordering.
Sub: docs/superpowers/specs/2026-05-27-recipe-cert-toggles-design.md
Task: T3.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Adds five requires_* Booleans on fusion.plating.process.node
(requires_coc, requires_thickness_report, requires_nadcap_cert,
requires_mill_test, requires_customer_specific), default True.
Recipe is SUPPRESS-ONLY: when False, the recipe never produces that
cert type even if the customer/part requested it. Default True =
existing recipes keep producing the same cert set they produce today.
Surfaced on recipe-level form (node_type == 'recipe'); resolver reads
from job.recipe_id which is always a top-level recipe node.
Post-migrate backfills NULL -> TRUE on existing nodes.
Sub: docs/superpowers/specs/2026-05-27-recipe-cert-toggles-design.md
Task: T2.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Adds three Boolean fields (x_fc_send_nadcap_cert, x_fc_send_mill_test,
x_fc_send_customer_specific) to res.partner, default False. Wires
aerospace/defence customers into the existing cert resolver so the
three orphan fp.certificate.certificate_type values become reachable.
Post-migrate idempotently backfills NULL -> FALSE on existing rows.
Sub: docs/superpowers/specs/2026-05-27-recipe-cert-toggles-design.md
Task: T1 of the implementation plan.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Adds recipe-level Boolean toggles (requires_coc / requires_thickness_report /
requires_nadcap_cert / requires_mill_test / requires_customer_specific,
default True) so a recipe can suppress certs the customer requested when
the recipe physically never produces them (passivation = no thickness,
commodity ENP = no nadcap).
Closes gaps on three orphan fp.certificate.certificate_type values
(Nadcap, Mill Test, Customer Specific) — adds partner toggles
(x_fc_send_nadcap_cert / x_fc_send_mill_test / x_fc_send_customer_specific,
default False), wires them through _resolve_required_cert_types, and
sets up manual-attach Issue flow (no QWeb auto-render for orphan types).
Brainstorming Q&A locked: recipe SUPPRESSES only, partner+recipe scope
(part-level unchanged), 5 booleans default True, manual PDF attach for
orphans.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Previous tightening removed the row-span but reintroduced a worse
problem: the tall PO block (with PO Pending + Expected By + chase
warning visible = ~250px) had only 2 small cells next to it
(Customer Job # / Job Sorting). 200px+ of vertical air below them
before row 3 started.
Layout now:
- Row 1: Customer (1-2) + Delivery Address (3-4)
- Rows 2-5 left: PO Block spans 4 grid rows (cols 1-2)
- Rows 2-5 right: 4 PAIRS of fields fill cols 3-4 in DOM order:
Row 2: Customer Job # + Job Sorting
Row 3: Material/Process + Lead Time
Row 4: Payment Terms + Delivery Method
Row 5: Pricelist + Quote Validity
- Row 6: Blanket SO + Invoice Strategy + conditional Deposit % / Progress %
(full 4-col width, kicks in after the PO block ends)
CSS Grid auto-flow places the right-side cells in the open positions
next to the row-span-4 PO block. Each grid row auto-sizes to the max
of the cells in that row (PO block top portion or the right pair),
so PO block height naturally aligns with the 4 right rows — no dead
air on either side.
User reported too much vertical air between fields. Two changes:
1. Removed grid-row: span 2 from the PO block. The row-span pattern
stretched each grid row to half the PO block's height (~125px each),
leaving empty space below Customer Job # / Job Sorting on row 2 and
below Material/Lead Time on row 3.
New layout:
- Row 1: Customer (1-2) + Delivery Address (3-4)
- Row 2: PO Block (1-2, naturally tall) + Customer Job # + Job Sorting
- Row 3: Material/Process + Lead Time + Payment Terms + Delivery Method
- Row 4: Pricelist + Quote Validity + Blanket SO + Invoice Strategy
- Row 5 (conditional): Deposit % or Progress % (when invoice strategy uses them)
PO block forces row 2 to be tall but cols 3-4 just sit at top — that
was the original mockup pattern, and it's denser overall because
rows 3+ are all the standard short height.
2. Tightened spacing in SCSS:
- Grid row gap 14px → 6px
- Cell label margin 0 (was 2px)
- Input padding 5px → 2px vertical, min-height 30px → 24px
- PO block padding 10px → 6/12/8px
- PO row gap 2px padding → 0 (min-height 28px keeps clickable target)
- PO chase text 11px → 10px, tighter line-height
My .o_fp_xpr_cell rule set width/height: 18px on every input[type=
checkbox], which broke Bootstrap's .form-switch slider proportions
(switches need width: 2em / height: 1em). Result: PO Pending and
other boolean_toggle widgets rendered as a single grey circle with
no visible track.
Excluded .o_field_boolean_toggle from the checkbox override and added
explicit Bootstrap form-switch styling — width: 2em, height: 1.2em,
accent colour on checked state, accent-bg focus ring. Non-switch
checkboxes (Blanket SO, Block partial shipments etc.) keep the 18px
square treatment.
H1 — Recipe propagation hardening for multi-part orders. The G3 onchange
fires when material_process changes, but a newly-added line (especially
via inline part create) sometimes didn't pick up the recipe before
confirm. In action_create_order, just BEFORE building so_vals, force
line.process_variant_id = wizard.material_process if the line is missing
one. Also added the same fallback inside the so_vals dict so the SO line
always carries the right recipe even if the wizard line missed it.
H2 — Strip 'spec - PART Rev X (xN)' header from customer-facing
description. Per user feedback, the customer-facing reports (SO
confirmation, Invoice, CoC, packing slip, BoL) should show ONLY the
typed description + thickness in the Description column. The legacy
header that prepended part metadata to line.name duplicated info from
the Part Number column. Wizard now writes ONLY the customer description
to line.name; the Part Number column owns the part-rev-name display.
H3 — Uppercase customer-facing description in reports. The shared
customer_line_description macro now wraps the description, serial,
and thickness in text-transform: uppercase divs. All reports that use
the macro (SO confirmation, Invoice, CoC, packing slip, BoL) get the
caps treatment automatically. Non-part lines (freight, rush fees)
keep their natural casing.
Manually cleaned up DOD-00154/SO-30062:
- Backfilled line 682 with the header recipe (ENP ALUM BASIC HP)
- Stripped the legacy 'No spec - PART Rev (xN)' header from both
lines' names; descriptions now read 'THIS IS TEST SPECIFICATIONS...'
and 'THIS IS BLB ABLA BOLL' cleanly.
Three cascading bugs caused DOD-00153/WO-30061 to confirm with zero
steps (and DOD-00150 to keep masking/bake even with overrides):
1. _is_node_included() in fp_job._generate_steps_from_recipe consulted
the per-job override_map ONLY when node.opt_in_out was 'opt_in' or
'opt_out'. Default is 'disabled' (mandatory), so overrides on
mandatory recipe nodes (Masking, De-Masking, Oven baking) were
silently ignored. Fix: consult override_map FIRST — explicit per-job
override always wins, regardless of node's opt_in_out value.
2. fp.direct.order.line.recipe_choice_ids didn't include the wizard's
material_process recipe (Express Orders order-level recipe), so the
line's process_variant_id domain rejected propagation. Added a 4th
tier to the compute that pulls the order's header recipe in.
3. sale_order._fp_resolve_recipe_for_line fell back from line picker
to part.default_process_id with nothing between. Added Express
header recipe (self.x_fc_material_process) as a 2nd-priority
fallback — catches cases where G3 propagation failed to reach the
line but the SO header has the recipe set.
Also fixed an unrelated G4 bug: _FP_PART_SYNC_FIELDS mapped
process_variant_id → 'default_process_variant_id' which doesn't
exist. Real field is 'default_process_id' (singular).
Cleaned up DOD-00153/WO-30061 manually: backfilled line +
job.recipe_id, regenerated steps with overrides respected. 8 steps
now visible, masking/bake correctly omitted.
Root cause: spec + plan assumed fusion.plating.process.node.default_kind
values for masking/baking nodes were 'masking', 'de_masking', 'baking'.
Actual values per inspection of WO-30060 / recipe ENP-ALUM-BASIC:
- 'mask' (Masking step)
- 'demask' (De-Masking step)
- 'bake' (Oven baking / Oven bake post de-rack)
So _fp_apply_express_overrides_to_job was searching for nodes that
don't exist → no override rows created → step generation included
masking + bake even when the SO line had x_fc_masking_enabled=False
and x_fc_bake_instructions=empty.
Fixed all 4 occurrences in _fp_apply_express_overrides_to_job:
- pre-deletion search uses ('mask','demask','bake')
- masking opt-out walker calls ('mask','demask')
- bake opt-out walker calls ('bake',)
- bake step instructions filter uses default_kind == 'bake'
Manually cleaned up DOD-00150 / WO-30060:
- Deleted the 4 masking/bake steps that were wrongly created
- Created 4 override rows so any re-generation respects the opt-outs
Future orders with masking off / bake empty will correctly skip these
recipe nodes at step-generation time.
Regression from G2 conversion (Char → Many2One). The wizard's
action_create_order built so_vals with 'x_fc_material_process':
self.material_process (the recordset) instead of .id. Passing a
recordset where an integer FK is expected raised:
psycopg2.ProgrammingError: can't adapt type 'fusion.plating.process.node'
at sale.order create time, breaking Confirm Order.
Python-only fix — no module upgrade needed, systemctl restart picks
it up.
Root cause: my @api.depends_context('fp_express_part_picker') decorator
on _compute_display_name was not honored by Odoo. Verified via odoo
shell — display_name returns the full 'PART (Rev X) — Name' regardless
of context. Reason: display_name is defined on the base Model class
and Odoo registers the field metadata (incl. _depends_context) when
the field is FIRST declared. Subclass redefinitions of the compute
method don't update _depends_context after the fact.
Workaround: don't rely on display_name context override. Instead,
overlay a custom span on top of the Many2OneField that shows JUST
the part_number_display value. CSS overlay uses:
- position: absolute / inset: 0
- background: $xpr-card (matches list row background)
- z-index: 2 over the picker
- pointer-events: none so clicks pass through to the picker
When the picker is focused (:focus-within parent), the overlay
hides so the user sees the autocomplete input value as they type.
When not focused, the overlay covers display_name with just the
part number.
Row 1 now reads 'ENG-1042 / B' — picker on the left (showing only
part_number_display), separator, revision on the right. Matches the
mockup pixel layout the user requested.
Customer feedback: rows 2 (description) and 3 (serials) in the Part
cell rendered as read-only spans. User wanted to edit directly.
New writable computed fields on fp.direct.order.line:
- part_name_editable: compute reads part_catalog_id.name, inverse
writes back to part.name on the linked catalog record
- serials_text: compute joins serial_ids names with commas; inverse
parses the typed string and find-or-creates fp.serial records,
updates the line's serial_ids M2M
Removed the redundant rev separator (display_name already includes
'(Rev X)' so showing it twice was clutter). Rev edits happen by
editing the part record directly via the OPEN button.
OWL widget templates updated:
- Row 2: <input> bound to part_name_editable, t-on-change saves
- Row 3: <input> bound to serials_text, t-on-change parses + saves
SCSS:
- Row 2 input: italic, transparent border, focus tints background yellow
- Row 3 input: small grey text, comma-separated friendly placeholder
- Both disabled-look when no part is picked
Both inputs trigger the inverse method on blur. The G4 sync chain
takes over from there to push line.line_description etc. back to
the part as before — so editing in the line keeps the part defaults
fresh for future orders.
Four customer-feedback fixes (G1-G4):
G1 — Part cell display redundancy. fp.part.catalog.display_name was
showing 'PART (Rev X) — Name' which duplicated with my Part cell widget's
separately-rendered revision + name rows. Added @api.depends_context
('fp_express_part_picker') to _compute_display_name: when the context
flag is True, display_name returns JUST the part_number. The Express
view passes the flag on the part_catalog_id field, so the picker shows
'9876699373' and the widget's row 2/3 show the rev + name.
G2 — Material/Process Tag is now the order's RECIPE, not a free-text
shop tag. Converted material_process from Char to Many2One(fusion.
plating.process.node) with domain [('node_type','=','recipe')] on both
fp.direct.order.wizard AND sale.order. Pre-migration (19.0.22.1.0/
pre-migrate.py) drops the old VARCHAR column so Odoo recreates as
INTEGER FK. Per dev-stage policy, old tag data is dropped.
G3 — Auto-apply order recipe to every line. New onchange
_onchange_material_process_apply_to_lines on the wizard: when the
header recipe is picked / changed, propagate to every line's
process_variant_id (unless the line has an explicit per-line override
that doesn't match the previous header value).
Plus an override on fp.direct.order.line.create that seeds new lines'
process_variant_id from wizard.material_process. So a newly-added
line auto-inherits the order's recipe.
G4 — Auto-sync line edits back to the part catalog. New
_fp_sync_to_part method called from create() + write() on
fp.direct.order.line. Tracked fields:
- line_description → part.default_specification_text
- bake_instructions → part.default_bake_instructions
- thickness_range → part.x_fc_default_thickness_range
- masking_enabled → part.default_masking_enabled
- process_variant_id → part.default_process_variant_id
Future orders for the same part will auto-pull these updated defaults
via the existing _onchange_part_default_thickness chain. Last-write-
wins semantics across concurrent edits (acceptable per dev-stage).
Wasted-space audit revealed:
1. PO Block occupied a wide 2-col span but its inner inputs didn't
fill the available width (130px label column + narrow input area)
2. Customer Job # + Job Sorting placed in row 2 cols 3-4 next to the
tall PO block left empty vertical space below them
3. Row 2 col 3-4 was sparse because the PO block forced row 2 to be tall
Reorganized:
- PO Block now spans 2 grid ROWS (.row-span-2 class → grid-row: span 2)
- Customer Job # / Job Sorting flow into row 2 cols 3-4 (alongside PO top)
- Material/Process Tag / Lead Time flow into row 3 cols 3-4 (alongside
PO bottom) — filling the previously-empty space next to the PO block
- Row 4 (after PO ends): Payment Terms / Delivery Method / Pricelist /
Quote Validity — full 4-col width back
- Row 5: Blanket Sales Order + Invoice Strategy + conditional Deposit % /
Progress Initial % (only show when relevant invoice strategy picked)
Inside the PO block:
- Label column tightened 130px → 110px so the input takes more width
- Inputs + Many2One wrappers now have width: 100% propagated, so PO #
and Expected By inputs fill the available row width
- Upload button restyled with the accent colour (was the green default)
Net effect: same field count but tighter packing, no empty vertical
or horizontal space next to the PO block.
Root cause for column widths: Odoo 19's column_width_hook.js dynamically
sets inline widths on every cell at render time, overriding any CSS
width on td/th selectors. Confirmed by reading the hook source on
entech: 'A width can also be hardcoded in the arch (width="60px").'
Fix: set width='Npx' as an ARCH ATTRIBUTE on each <field> in the line
list:
- Part Number 230px, Line Job # 80px, Thickness 100px, Mask 55px,
Bake 120px, Qty 55px, Price 80px, Subtotal 90px, Action stack 60px
- Specification + Internal Notes get NO width → take remaining flex
space (responsive: layout adapts to viewport)
Root cause for missing checkbox: my SCSS underline-style override
selected ALL .o_field_widget input including type=checkbox, rendering
checkboxes as 30px-tall full-width transparent text inputs.
Fix: exclude type=checkbox/radio/file from the underline rule, and
add explicit rendering for type=checkbox (18px square, accent-coloured)
inside .o_fp_xpr_cell. The Blanket Sales Order checkbox + the inline
Block partial shipments checkbox are now both visible.
1. Blanket Sales Order — match legacy field shape. Renamed label from
'Blanket SO' to 'Blanket Sales Order' (matches legacy view), removed
the boolean_toggle widget (defaults to checkbox), and added the
sibling 'block_partial_shipments' field inline (only visible when
blanket is checked, with 'Block partial shipments' helper text).
2. Column widths — give roomier columns where data needs space, tighten
numeric columns. Part Number 230px, Specification min 220px,
Internal Notes min 140px, Qty 60px, Price 80px, Subtotal 90px,
Mask 55px, Bake 130px, Action stack 60px.
3. Stacked DWG / OPEN buttons — new OWL widget FpExpressActionBtns
(express_action_btns.js + .xml) renders both buttons vertically in
ONE column to save horizontal space. Widget binds to a new
action_btns_anchor field (related from part_catalog_id) on the
line. Each button shows tooltip + disabled state when no part is
picked; DWG triggers the native file picker, OPEN navigates to the
part record.
4. Field activation — clicking the cell anywhere now focuses the
input, not just clicking the label. Achieved via:
- cursor: text on .o_fp_xpr_cell
- cursor: pointer on labels
- min-height: 30px on all inputs (larger click target)
- width: 100% propagated through Many2One wrappers (.o-dropdown,
.o-autocomplete) so the input genuinely fills the cell
- box-sizing: border-box so widths are predictable
- Background tint on focus for visual feedback
A bad replacement in the previous commit left an extra '}' that
prematurely closed the .o_fp_xpr block, dumping all the legend bar /
PO status pill / part cell / bake pill styles OUTSIDE the namespace.
SCSS compile silently produced an unusable bundle and the form
rendered without any of the new visual treatment.
Brace balance now verified at 0.
NEW OWL widgets:
- FpExpressPartCell (static/src/js/express_part_cell.js + .xml) — multi-row
Part cell. Wraps Many2OneField for the part picker (row 1: part# / rev,
bold). Below it: row 2 part description (italic muted), row 3 serial #s
joined + '+ bulk' button that triggers the existing bulk-add wizard.
- FpExpressBakePill (static/src/js/express_bake_pill.js + .xml) — click-
to-edit Bake pill. Renders amber pill when set, italic muted 'no bake'
when empty. Click swaps to inline textarea + Save / Clear / Cancel.
NEW fields:
- fp.direct.order.line.part_number_display / part_revision_display /
part_name_display (related Char from fp.part.catalog) — fed to the
Part cell widget so it can render multi-row without RPC.
- fp.direct.order.wizard.tooling_charge (Monetary) — surfaced in the
Totals card on the Express form.
- fp.direct.order.wizard.po_status (computed Selection
received/pending/missing) — drives the PO Block status badge.
- sale.order.x_fc_tooling_charge (Monetary) — receives wizard.tooling_charge
at confirm.
View updates (fp_express_order_views.xml):
- PO block header now shows the PO status pill (green Received,
amber Pending, red Missing)
- Order Lines legend bar (Mask / Bake pill / DWG / OPEN explainers)
- Part Number column uses widget='fp_express_part_cell' — single cell
with 3 internal rows
- Bake column uses widget='fp_express_bake_pill' — interactive pill
- Totals card now has Subtotal / Tooling Charge / Total Lines / Total
Quantity / Grand Total + currency pill
SCSS adds:
- Multi-row part cell styles (internal borders, bold/italic/muted rows)
- Bake pill (has-bake amber, no-bake italic muted) + inline editor
- Legend bar (section background, gap-spaced explainer chips)
- PO status pill colour scheme
- Bulk button styling
Wizard's action_create_order now carries tooling_charge to the SO at
confirm so it persists on the resulting sale.order.
Previous rebuild used Odoo's <group col='4'> which renders as an HTML
table — colspan+nesting broke into a vertical stack. Replaced entirely
with raw <div> + CSS Grid (display: grid; grid-template-columns:
repeat(4, 1fr)) so the header layout matches the mockup exactly:
- Row 1: Customer (span 2) + Shipping Address (span 2)
- Row 2: PO block (span 2, accent-bordered card with PO#/PDF/Pending
toggle/Expected date stacked + chase warning) + Customer Job # + Job Sorting
- Row 3: Material/Process Tag + Lead Time (inline X to Y) + Payment
Terms + Delivery Method
- Row 4: Blanket SO + Currency/Pricelist + Quote Validity + Invoice Strategy
Footer also rebuilt as CSS Grid (1fr 320px) — Notes/Terms cards
stacked in the left column, Totals card with Grand Total + currency
pill in the right column. Each card has a title + subtitle + body
matching the mockup's card chrome.
SCSS overrides Odoo's default field chrome inside .o_fp_xpr_cell so
inputs render with the mockup's underline style (no Bootstrap form-
control border, just a 1px bottom-border that thickens on focus).
Restructures the Express form to align with the brainstorming mockup:
Header (4-column grid via <group col='4'>):
- Row 1: Customer (colspan=2) + Shipping Address (colspan=2)
- Row 2: Consolidated PO Block (colspan=2 with PO#/PDF/Pending toggle/
Expected date stacked + chase warning inline) + Customer Job # + Job Sorting
- Row 3: Material/Process Tag + Lead Time (X to Y inline) + Payment Terms + Delivery Method
- Row 4: Blanket SO + Currency/Pricelist + Quote Validity + Invoice Strategy
Lines: 13 inline columns including the Express-specific Line Job #,
masking toggle, bake text, plus per-line action buttons (DWG, OPEN,
+ bulk) wired to the Phase B helpers.
Footer: side-by-side cards — Notes + Terms stacked in the left card,
Totals card on the right with Total Lines / Total Qty / Grand Total
+ currency pill.
SCSS adds:
- PO block: accent-bordered card-within-card
- Lines: tight spreadsheet borders, hover row highlight
- Bake column: amber pill style, italic 'no bake' for empty
- Customer Line Job #: bold, uppercase, narrow column
- Inline action buttons: small uppercase bordered chips
- Footer cards with prominent Grand Total + currency pill
OWL multi-row Part cell (FpExpressPartCell) and click-to-edit Bake
pill (FpExpressBakePill) are still deferred — they need real OWL
components, separate pass.
C1: product.pricelist._compute_display_name override gated by the
'fp_express_currency_picker' context flag (set on the Express form's
pricelist_id field). When active, prefixes the dropdown label with
the currency code: 'CAD — Public Pricelist (CAD)'. Elsewhere the
standard display name is unchanged.
C3: SCSS tokens + base styles for the Express form. Tokens use the
compile-time @if $o-webclient-color-scheme branch per the project's
'Dark Mode' rule — same SCSS compiles into both bundles with different
hex values. Token vars wrapped in CSS custom properties so downstream
modules can override for per-shop branding without recompiling.
Base styles: spreadsheet-feel table borders, bake-cell inset-pill,
customer line ref bold/uppercase, accent section markers.
D1: action_open_draft method routes drafts-list click to the matching
form view by view_source. view_source badge column added to the drafts
list (Express=blue, Legacy=muted).
D2: Deprecation banner on the legacy direct-order form pointing operators
to the new Express view, plus a Switch-to-Express header button. Legacy
action context defaults new drafts to view_source='legacy' (Express
action defaults to 'express') so newly-created drafts open in the right
view automatically.
Consolidates the brainstorming session into a single design spec. Covers:
header layout + field-to-model mapping, line widget (multi-row Part cell,
masking + bake pills, serial bulk-add trigger), masking/baking override
flow at SO confirm, currency/pricelist picker mechanics, inline part
create + drawing upload + open-part buttons, and phase-out path for the
legacy Direct Order view.
Reuses fp.direct.order.wizard model end-to-end (Q1=D). New fields:
material_process, customer_line_ref, masking_enabled, bake_instructions,
default_specification_text, default_bake_instructions,
default_masking_enabled, x_fc_internal_notes, x_fc_print_terms.
Rename: wizard.notes → terms_and_conditions. Retire: wizard.currency_id.
Mockup at .claude/mockups/express_orders.html (interactive, light + dark).
User is on Mac via Tailscale into this Windows host. Browser previews
bound to Windows localhost are unreachable from the Mac browser. Default
to text-based design discussion on this host instead of spinning up the
brainstorming visual companion. Has bitten three times now.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Captures all clarifying-question answers + exploration findings so a
fresh Claude Code session on Mac can resume at 'propose architectural
approaches' without re-running the discovery work.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Two bugs caught by entech battle test on first deploy:
1. Manifest entry landed in the 'demo' list instead of 'data' because
my anchor (fp_demo_shopfloor_data.xml) was already in 'demo' —
the entry pattern-matched into the wrong section. Demo data
doesn't load on entech (no --load demo), so the mail.template
never existed. Moved fp_tablet_pin_reset_template.xml to 'data'.
2. The fp.notification.template wrapper record referenced a model
that doesn't exist until fusion_plating_notifications loads;
fusion_plating_shopfloor doesn't depend on notifications, so
the data load ParseError'd. Removed the wrapper — the controller
calls mail_template.send_mail() directly anyway, not via the
notification dispatcher. Added an inline comment explaining why
the wrapper isn't here.
Battle test updated to drop the (now removed) wrapper xmlid check.
Battle test ALL PASS on entech after fixes.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
4 new state-machine modes on FpTabletLock, reusing the existing
FpPinPad 4-cell component:
- request_code : 'Send temp PIN' button screen (no-PIN tile OR
after 3-fail Forgot button)
- enter_temp_code : 4-cell pad for the emailed code
- set_new_pin : 4-cell pad — choose new PIN
- confirm_new_pin : 4-cell pad — confirm new PIN
Trigger paths (per D1 + D2):
- Tap no-PIN tile -> goes straight to request_code mode
(onTileClick dispatches via tile.has_pin)
- Wrong PIN 3 times -> 'Forgot? Reset PIN via email' button appears
below the pad (gated by state.failedAttempts >= 3)
Client-side failedAttempts counter (resets on tile re-select per D14).
Server-side x_fc_tablet_pin_failed_count keeps incrementing to the
existing 5-fail lockout per D13.
After Confirm New PIN succeeds, auto-login fires unlock_session with
the new PIN. If unlock_session fails for any reason, falls back to
'PIN set, tap your tile to log in.' status.
SCSS reuses $lock-* tokens from _tablet_lock_tokens.scss — light +
dark handled by the existing token system (no new tokens needed).
Hand-Off gold gradient repeated for the primary 'Send temporary PIN'
button to match the existing tablet visual language.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Mail template renders the 4-digit code in both subject (mobile
notification glance) and body (big bold display). Per Rule 25 only
core res.users fields referenced; the code itself comes from ctx.
fp.notification.template wrapper enables admin UI customization of
the body without touching code. tablet_pin_reset_requested added to
TRIGGER_EVENTS selection.
Daily ir.cron purges used/expired rows > 7 days old (audit trail
lives in fp.tablet.session.event, not here, so aggressive cleanup
is safe).
Manifest bump 19.0.34.2.0 -> 19.0.35.0.0 (triggers asset cache
invalidation on -u so the new template + SCSS load cleanly).
Phase 1 backend complete.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Three controller changes in one commit (tight code coupling):
1. /fp/tablet/request_reset_code (Task 2) — generates 4-digit code,
emails it, returns masked_email. Specific error codes for the
frontend to switch on (no_email + manager_name, rate_limited +
wait_minutes, user_not_found, no_role, inactive). Shop-branch
role check matches existing _check_credentials per Rule 13l + 23
(all_group_ids transitive — Owners reach Technician through
implication).
2. /fp/tablet/verify_reset_code (Task 3) — verifies the emailed
code, on success mints a 5-min HMAC reset_token. Error responses
are specific (no_active_code / expired / too_many_attempts /
wrong_code with attempts_left).
3. set_pin extended to accept reset_token (Task 4) — three auth
paths now: old_pin (existing), reset_token (new), or neither
(existing — only for users with no current hash). reset_token
path is the only one that operates on a user OTHER than env.user;
token proves the legit user just verified their email.
Failure audit reuses existing failed_unlock event_type with a notes
field describing the reset-code-specific reason. Success audit uses
the new pin_reset_requested / pin_reset_code_verified /
pin_set_after_reset event_type values.
_mask_email helper added for the no-email-on-file edge case.
3 more tests cover: valid token roundtrip + set_pin, expired token
rejection, and lockout-cleared-on-reset.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
fp.tablet.pin.reset stores hashed 4-digit codes emailed for self-
service PIN create/reset. Per CLAUDE.md Rule 24 + Rule 13l it follows
the defensive patterns established elsewhere in the shopfloor module:
- PBKDF2-SHA256 hashing (200k iterations, matches ResUsers PIN)
- 72h TTL per D4
- 5 wrong-attempt cap per D5 (invalidates code, used_at set)
- 3 requests/60min rate limit per D6 (raises UserError)
- SQL EXCLUDE constraint enforces one-active-row-per-user per D7
- HMAC-SHA256 reset_token (300s TTL, single-use) for step 3 of
the flow (set_pin via reset_token alternative to old_pin)
Audit event_type extended with 3 new values (pin_reset_requested,
pin_reset_code_verified, pin_set_after_reset). Manager-only ACL on
the new model; sudo when endpoints need access.
10 model-level tests cover generate / replace-active / rate-limit /
verify-correct / verify-wrong / 5-attempt-cap / expired / token sign
roundtrip / tampered-sig / purpose-mismatch.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
User goal: from the Shop Floor Terminal lock screen, a user with no
PIN (or who forgot their PIN) should be able to set / reset their
own PIN without a manager's help. Today, FpPinSetup runs only from
Preferences which requires being logged in — there's no path from
the lock screen.
Design (approved, with user-picked defaults):
- Tap tile of no-PIN user -> 'Send temporary PIN' button -> email
4-digit code, valid 72 hours -> enter code -> choose new PIN ->
auto-login.
- For existing-PIN users: 3 failed PIN entries -> 'Forgot? Reset
PIN via email' button appears below keypad -> same email flow.
- Both flows merge at: enter temp code -> set new PIN.
- Email goes to res.users.login (or partner_id.email fallback).
No-email-on-file -> 'Contact your manager: <owner>' message.
- Rate limit: 3 requests per user per rolling 60 min.
- Per-code cap: 5 wrong attempts invalidates the code.
- New model fp.tablet.pin.reset stores hashed code + expires_at
with SQL constraint enforcing one-active-row-per-user.
- 2 new endpoints (request_reset_code, verify_reset_code) + extend
existing /fp/tablet/set_pin to accept reset_token alternative
to old_pin.
- Audit: 3 new event_type values on fp.tablet.session.event.
- Reuses existing PBKDF2 helpers, FpPinPad component (mode prop),
fp.notification.template dispatch, mail.template pattern.
Per CLAUDE.md Rule 25 the mail template references ONLY core
res.users fields (object.name, object.email, object.login,
object.company_id) — ctx.code is dispatched as extra_context, not
a model field. Safe at parse-time.
Self-review fixed 2 issues:
- event_kind -> event_type (real field name on fp.tablet.session.event)
- Listed existing event_type values explicitly for context
Spec: docs/superpowers/specs/2026-05-25-tablet-pin-self-service-design.md
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Plan-time xmlids were wrong — entech battle test caught all 5
non-cert kanban xmlids missing. Real xmlids (queried via
ir.model.data on entech):
hold: action_fp_quality_hold (was action_fusion_plating_quality_hold)
ncr: action_fp_ncr (was action_fusion_plating_ncr)
rma: action_fp_rma (was action_fusion_plating_rma)
capa: action_fp_capa (was action_fusion_plating_capa)
check: action_fp_quality_check (was action_fusion_plating_quality_check)
cert stays unchanged — action_fp_certificate was already correct.
After fix: battle test ALL PASS — 6 sections in canonical order,
all xmlids resolve, 3 banner items pulled from real entech data
(5 draft certs, 3 of them overdue past 24h).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Version 19.0.7.0.0 → 19.0.8.0.0 (triggers asset cache invalidation
on -u so the new template + SCSS load cleanly).
Battle test script: 6-check entech smoke. Validates snapshot shape,
canonical section order, required section keys, open_kanban_xmlid
resolves to act_window, banner item shape when items exist. Summary
prints per-section counts so you can eyeball the entech state.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
JS: single FpQualityDashboard component + BannerCard / BannerItem /
SectionCard / SectionRow sibling sub-components in the same file.
Fetches /fp/quality/dashboard/snapshot, 60s poll, deep-link
?tab=certificates scrolls to section-cert via scrollIntoView.
XML: outer wrapper + banner + 6 sections (t-foreach over
state.snapshot.sections). Each section has id='section-<type>' so
the deep-link target works. SectionRow has overdue-conditional
class for red subtitle highlight.
SCSS: local tokens for urgent/good/section-head with light+dark via
$o-webclient-color-scheme branch. 135deg gradients matching the
plant kanban polish. Mobile breakpoint at 900px collapses banner
grid to 1 col and stacks row Open button.
OLD TABS array, selectTab, openTab, totalOpen, totalOverdue all
deleted. Old template's tab tiles + per-tab panels deleted. Existing
per-model kanbans untouched.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Covers: missing-field critical-customer check returns empty without
crashing; computed_at is a valid ISO timestamp; every section ships
a non-empty open_kanban_xmlid in module.xmlid format.
(missing-model test from the plan omitted — patching env.__contains__
was unsafe; the in-self.env guard is already exercised by Tasks 2-4
in production behavior. The other 3 defensive tests still cover the
missing-field path, which is the more common scenario.)
Phase 1 backend complete.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
_fetch_banner_candidates collects (overdue) OR (critical-customer +
open) records per type. _critical_customer_ids reuses partner.x_fc_rush
and partner.x_fc_vip flags when defined (gracefully no-ops when
absent). _critical_badge returns RUSH/VIP/AEROSPACE/AS9100 label
when the banner reason is critical-customer (no badge when overdue).
_build_banner ranks: overdue first by oldest, then critical-customer
by oldest, takes top 6, reports total_matching.
build() now collects banner candidates from every section in one
pass + invokes _build_banner once.
Tests cover overdue hold pickup, 6-cap with overflow count, and
all_clear when DB is empty.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
_fetch_section_items pulls top-5 open records per type, ranked
overdue-first by oldest create_date. _build_item shapes each row
with id/name/customer/subtitle/urgency/open_action. _resolve_partner
defensively walks partner_id -> job_id.partner_id -> ncr_id.partner_id
per type. _build_subtitle generates the human-readable second line.
Tests cover empty list, 5-cap on 8-record set, and required item
keys (id/name/customer/subtitle/urgency/open_action).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Replaces /counts with /snapshot. Helper class FpQualityDashboardSnapshot
returns response with correct shape — banner placeholder + per-type
sections with open/overdue counts (reuses old counts endpoint
thresholds). Items + critical-customer banner come in Tasks 3-5.
Per CLAUDE.md Rule 13m, Model.sudo() on cross-module reads. Per
Rule 24 the in-self.env check guards missing-model paths.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Tests for empty-DB all-clear, canonical section order, and required
keys on each section. All fail until Task 2 lands the snapshot helper.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
User goal: 'all quality related updates at glance, all the flagged
tasks need to show right here so the manager can quickly follow up
and complete the task'. Current dashboard is a tab-router (6 numeric
tiles + click-to-drill) — flagged tasks aren't visible without
navigation.
Design (Hybrid layout, approved):
- Red 'Needs Attention Today' banner on top (up to 6 items, 2x3 grid)
showing items that are overdue OR from critical customers
(x_fc_rush / x_fc_vip / aerospace). Green 'all caught up' when zero.
- Per-type sections below in QM-urgency order: Certs / Holds / NCRs /
RMAs / CAPAs / Checks. Each shows top 5 items inline + Open all
link to the existing kanban.
- Single 'Open ->' button per row -> opens record form via act_window.
No one-click action shortcuts (cert form is where Fischerscope +
sign-off prereqs are validated).
- Drop the existing 'Quality Overview' header strip entirely.
- 60s poll cadence preserved.
- ?tab=certificates deep-link from awaiting-cert notification email
preserved as scrollIntoView on the certs section.
Backend: replace /fp/quality/dashboard/counts with /snapshot. New
helper class FpQualityDashboardSnapshot builds banner + 6 sections in
one response. Cross-module reads sudo'd per Rule 13m; missing fields
gracefully degrade per Rule 13j defensive pattern.
Frontend: rewrite the OWL component. BannerCard + 6 SectionCards as
sub-components in the same JS file (not reused elsewhere yet).
Existing per-model kanbans untouched.
Self-review fixed 4 issues:
- _critical_customer_domain made per-type (was contradictory)
- OVERDUE_THRESHOLDS gained explicit use_due_date flag (CAPA branch)
- Template requirement called out: id='section-<type>' on each card
for the deep-link scrollIntoView to work
- doAction call shape disambiguated for xmlid vs full dict
Spec: docs/superpowers/specs/2026-05-25-quality-dashboard-redesign-design.md
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
User feedback after first deploy: 7 KPI tiles wrapped to second line
(grid was repeat(5, 1fr) but I had added 2 new ones), and the
controls felt cramped.
Layout fix:
- .kpi-strip grid: repeat(5, 1fr) → repeat(8, 1fr) so the row stays
one line and there's room for the new Awaiting QC tile.
Missing KPI added:
- Awaiting QC — fp.job.card_state='awaiting_qc' count. Operators
couldn't see when QC was blocking job close from the KPI strip
(only visible inside the column). Server-side count + filter
clause + matching filter chip.
Visual polish (all light + dark via existing token system):
- KPI tiles: padding 6→10px, value font 20→26px, label font 9→10px,
subtle 135deg linear-gradient bg per kind (urgent/warn/good/qc),
hover lifts the tile with translateY + shadow.
- Filter chips: padding 4/12→7/16px, font 11→13px, gradient bg,
active state has gradient blue + shadow.
- Search input: padding 5/10→9/14px, font 12→14px, focus ring.
- Toolbar buttons (Station/All Plant/Manager/Scan QR/Hand Off):
padding 5/10→8/14px, font 12→14px, gradients, hover lift.
Dark mode handled automatically — all gradients reference
$plant-* tokens which already have @if $o-webclient-color-scheme ==
dark global overrides in _plant_tokens.scss.
Version bump fusion_plating_shopfloor 19.0.34.0.0 → 19.0.34.1.0.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Three bugs caught + fixed during entech battle test:
1. _fp_check_finish_gates calling button_mark_done triggered the
step-completion gate prematurely (step still in_progress at
pre-super time). Pass fp_skip_step_gate=True alongside
fp_check_gates_only — we know the operator is about to finish
the last open step.
2. _fp_schedule_cert_activity used env.get('fp.notification.template')
for presence check. env.get returns an EMPTY recordset (falsy),
not None — 'if not Template: return' silently exited and no
activity was ever scheduled. Switch to 'in self.env' check
pattern + explicit indexing. CLAUDE.md Rule 24.
3. _fp_check_advance_after_cert_issue + _fp_check_regress_after_cert_void
used 'state != issued' as outstanding-cert count. This made
voided certs count as outstanding forever, so void+re-issue
cycles never re-advanced. Switch to per-type coverage check:
each required cert TYPE needs at least one issued cert.
Regress mirrors: only fire if a type loses all issued certs.
CLAUDE.md gains Rule 24 (env.get falsy empty recordset trap).
Rule 25 (mail.template parse-time validation) renumbered.
Battle test ALL PASS on entech admin DB:
10/10 steps green — auto-advance, kanban placement, activity
schedule + auto-resolve, ACL guard, cert issue advance, void
regress, re-issue advance, manual ship.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Entech deploy of 5a039ae3 hit:
ParseError: Failed to render inline_template template
AttributeError('fp.job' object has no attribute 'display_wo_name')
Root cause: mail.template data files are parse-time validated by
Odoo (template rendered against sample object). fusion_plating_notifications
loads BEFORE fusion_plating_jobs in dep order, so jobs-module fields
(display_wo_name, part_catalog_id) aren't on the Python class yet
even though the DB columns exist from previous installs.
Fix: strip display_wo_name → name and remove the Part row.
Recipe / qty_done / partner_id stay (all in fusion_plating core).
Logged as CLAUDE.md Rule #24 — same trap will bite anyone else
adding cross-module mail templates. Includes structural alternatives
for callers that really need downstream fields.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
10-step battle test covering: auto-advance on last step finish,
kanban placement, QM activity, ACL guard, cert issue advance,
activity auto-resolve, cert void regress, re-issue, manual ship.
Tolerant of partial state — branches around the awaiting_cert path
when partner doesn't require certs (uses awaiting_ship path instead),
SKIPs subsequent steps when prerequisites fail, rolls back at end so
the DB stays clean.
Run on entech via odoo-shell after deploy.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Idempotent post-migrate that moves mid-flight in_progress jobs whose
recipe steps are all terminal into the appropriate new state:
- draft cert exists → awaiting_cert
- no cert required → awaiting_ship
done jobs left alone (historically completed, already shipped).
Card_state + mini_timeline_json recomputed for affected rows so the
plant kanban renders correctly on first page load.
Version bump 19.0.10.31.0 → 19.0.11.0.0 triggers the migration on -u.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
TRIGGER_EVENTS extended with three new events:
- cert_awaiting_issuance — fires on in_progress → awaiting_cert
- cert_voided_re_notify — fires on awaiting_ship → awaiting_cert
regress (cert voided post-issue)
- job_shipped — fires on button_mark_shipped
_dispatch routes cert events through new internal-recipient resolver
(QM/Manager/Owner via all_group_ids, transitive per Rule 13l)
instead of the partner-based stream lookup. Other events unchanged.
Mail templates (fp_cert_authority_templates.xml): two new
mail.template records bound to fp.job. Amber accent bar for awaiting,
red accent bar for void-re-issue. Deep-link to
/odoo/action-...?tab=certificates so QM lands on the right tab.
Activity type (fp_activity_types_data.xml): mail.activity.type
activity_type_issue_coc — bound to fp.job, 1-day delay, certificate
icon.
fp.job helpers:
_fp_schedule_cert_activity: round-robin by oldest login_date,
idempotent on existing open activity, soft-fails if helpers
are missing.
_fp_resolve_cert_activities: auto-resolves on awaiting_ship,
soft-fails on per-activity exceptions.
Manifest bumps:
fusion_plating_notifications 19.0.6.6.1 → 19.0.7.0.0
fusion_plating_jobs: data list gains fp_activity_types_data.xml
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Counts endpoint: certificates block — open=draft, overdue=draft+>24h.
Falls back to {open:0, overdue:0} when fp.certificate isn't installed.
JS: TABS array gains the 6th entry. Existing data-driven OWL template
auto-renders both the header tile and the body panel. Tab opens the
fp.certificate kanban grouped by state, filtered to draft by default.
Deep-link: setup() reads action.context.params.tab. The
cert_awaiting_issuance notification email links to
/odoo/action-fp_quality_dashboard?tab=certificates and lands the QM
on the right tab automatically.
Template: 'Open across all 5' → 'Open across all <tabs.length>' so
it stays correct if more tabs are added later.
Manifest: fusion_plating_quality 19.0.6.6.6 → 19.0.7.0.0
(fusion_plating_certificates already in depends — no change needed).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Controller (plant_kanban.py):
- Widen domain: state IN (confirmed, in_progress, awaiting_cert,
awaiting_ship). Done jobs still drop off.
- _resolve_card_area: state=awaiting_cert → 'inspection' column,
state=awaiting_ship → 'shipping' column. State drives column
regardless of recipe shape.
- _state_chip: 🏷️ Awaiting CoC (amber) + 📦 Ready to ship (green).
- _SORT_PRIORITY: awaiting_cert=3.5, awaiting_ship=8.5.
- KPI dict: awaiting_cert + awaiting_ship counts.
- Filter clauses for the two new chips.
Model (fp_job.py):
- _compute_card_state handles new states in BOTH branches: the
no-active-step early return (where awaiting_cert/ship cards
land — all steps terminal) AND the per-step branch (defensive).
- _compute_mini_timeline_json: awaiting_cert paints inspection
dot 'current'; awaiting_ship paints shipping dot 'current'.
All earlier dots show 'done'.
SCSS (_plant_tokens.scss + _plant_card.scss):
- New tokens for amber (cert) + green (ship), light + dark variants
via the existing $o-webclient-color-scheme compile-time branch.
- .state-awaiting_cert / .state-awaiting_ship modifier classes
match the existing border-left pattern.
XML (plant_kanban.xml):
- Two new KPI tiles + two new filter chips wired to the state
filter clauses.
Manifest: fusion_plating_shopfloor 19.0.33.2.0 → 19.0.34.0.0.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
action_issue gated to Manager/QM/Owner via Python AccessError +
view-level groups= on the Issue button (two-layer enforcement).
Manager bypass via fp_skip_cert_authority_gate=True context flag
with chatter audit.
action_issue post-callback calls job._fp_check_advance_after_cert_issue
so the job auto-advances awaiting_cert → awaiting_ship when every
required cert is issued.
write({'state':'voided'}) override calls
job._fp_check_regress_after_cert_void so a previously-issued cert
being voided slides the job back to awaiting_cert and re-notifies
the QM.
x_fc_age_hours non-stored Float drives the Quality Dashboard age
chip + overdue filter.
Version bump 19.0.7.9.3 → 19.0.8.0.0 (spec said 19.0.6.0.0 but
current is already higher; bumped to next major instead).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
button_mark_shipped: manual transition awaiting_ship → done. Does
not re-run the bake/qty/QC gates — those passed at the in_progress
→ awaiting_cert/ship transition. Just the 'yes, shipped' stamp.
Milestone cascade (_compute_next_milestone_action) extended to
recognize the two new states:
- awaiting_cert → 'issue_certs' button
- awaiting_ship → 'mark_shipped' button
Legacy state='done' branch preserved for historical jobs.
action_advance_next_milestone now dispatches 'mark_shipped' via
_action_mark_shipped_dispatch which routes:
awaiting_ship → button_mark_shipped (new path)
done + active delivery → _action_mark_active_delivery_delivered
(legacy, unchanged)
View: 'Mark Shipped' milestone button gated on Manager/Owner groups.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Pre-super: when finishing the last open step on an in_progress job,
run the bake/qty/QC gates from button_mark_done so failures surface
as UserError on the click (per spec D12). Without this the
auto-advance would silently fail with no error path.
Post-super: trigger _fp_check_advance_post_shop so the state
auto-advances cleanly (in_progress → awaiting_cert / awaiting_ship).
Added _fp_check_finish_gates helper on fp.job and a
fp_check_gates_only context flag honored by button_mark_done so the
gate logic is single-sourced (DRY).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
_fp_check_advance_post_shop: in_progress + all steps terminal →
awaiting_cert (cert required) or awaiting_ship. Auto-spawns cert
+ delivery and fires notifications. Idempotent. Does NOT raise —
gate failures bubble up via fp.job.step.button_finish (Task 4).
_fp_check_advance_after_cert_issue: awaiting_cert → awaiting_ship
when every required cert is state=issued.
_fp_check_regress_after_cert_void: awaiting_ship → awaiting_cert
when a previously-issued cert is voided. Re-notifies QM.
hasattr guards on _fp_schedule_cert_activity + _fp_resolve_cert_activities
keep this safe during incremental rollout — those land in Task 20.
Test scaffolding added covering helper existence + idempotency.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Per spec docs/superpowers/specs/2026-05-25-post-shop-cert-shipping-job-states-design.md.
Selection extension only; transitions wired in subsequent tasks.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Trigger: WO-30058 (SO-30058) finished all recipe steps on entech and
disappeared from the Shop Floor kanban with a draft CoC and no QM
notification. Operators reported jobs feeling lost; risk that a job
could leave the building without paperwork.
Design (Approach A, approved):
- Two new fp.job.state values between in_progress and done:
awaiting_cert + awaiting_ship
- Auto-advance on last step finish; auto-advance on cert issue
- Plant kanban widens domain, renders the two states in the existing
Final inspection / Shipping columns
- 6th tab 'Certificates' on Quality Dashboard with kanban + filters
- ACL gate on fp.certificate.action_issue restricted to
Manager / QM / Owner (transitive via all_group_ids)
- Email + mail.activity notification to QM authority group
- Migration script backfills mid-flight jobs
Shipping label printing, BoL, carrier dispatch are explicitly
out of scope; awaiting_ship is a parking column with a manual
Mark Shipped button.
Self-review pass found and fixed:
- round-robin field ambiguity (last_activity_at vs login_date)
- unstated behavior for button_mark_done gates (now in step.finish)
- placeholder version inlined (19.0.11.0.0)
- dead reference replaced with inline body
Spec: docs/superpowers/specs/2026-05-25-post-shop-cert-shipping-job-states-design.md
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Two layout polish fixes after persona-walk feedback on the new Plant
Kanban surface (`fp_plant_kanban`).
1. Columns now run full board height with visible borders
Was: `.col` had `background: $plant-bg` (= page bg, invisible) and
no border, so only the header card (`.o_fp_col_header`) drew any
outline. Empty columns (BAKING / DE-RACKING / SHIPPING) looked
unbounded — operators couldn't tell where one column ended and
another began.
Now: `.col` is the bordered white card (Trello / Asana style),
stretches full height via grid + flex. `.o_fp_col_header` drops
its standalone border / radius / background and is just a bottom-
divider band inside the column card.
2. Horizontal scrollbar pinned to viewport bottom
Was: `.o_fp_plant_kanban` was `min-height: 100vh` (block flow) +
`.board` had `min-height: 520px; overflow-x: auto`. Scrollbar
showed at the bottom of the .board element (~520px from top of
board), floating mid-page below the empty columns.
Now: parent is `height: 100vh; display: flex; flex-direction:
column`. Header is `flex: 0 0 auto`; `.board` is `flex: 1 1 auto;
min-height: 0` so it fills all remaining vertical and its
scrollbar sits at the viewport bottom.
`.col-scroll` switched from `max-height: calc(100vh - 260px)` to
`flex: 1 1 auto; min-height: 0` so it expands inside the now-full-
height column instead of being capped at a magic number.
Version: fusion_plating_shopfloor 19.0.33.1.13.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Four fixes shipped together — all surfaced during tablet UX walkthrough
on entech.
1. sale.order ACL on step completion
Technicians hit "Access Denied... sale.order" when starting/finishing
any step. _fp_check_receiving_gate + the serial-promotion helpers +
_fp_resolve_contract_review_part read step.job_id.sale_order_id (and
sale_order_line_ids) without sudo. Per Rule 13m, denormalized cross-
module reads in tablet controllers must sudo() the source recordset.
2. Workspace stuck on "Loading Job Workspace…" after Hand Off + relogin
Action params aren't URL-encoded, so the workspace remounts with
jobId=null. refresh() exited early, state.data stayed null, "Loading"
shown indefinitely. onMounted now redirects to the plant kanban
when jobId is null or the initial load returns no data.
3. 4-hour timer offset on active steps
workspace_controller used fp_format() to serialize date_started —
which converts naive UTC to user tz wall time first. JS then
appended 'Z' and parsed as UTC, so timer was offset by the user's
tz (4h on EDT). Switched to fp_isoformat_utc() (proper +00:00 ISO)
and dropped the Z-append in formatActiveStepElapsed +
isActiveStepOvertime.
4. Lock-screen clock follows FP regional setting
tablet_lock.js used d.getHours() / d.toLocaleDateString() — browser
tz. Now /fp/tablet/tiles returns tz_name (fp_user_tz resolution:
user.tz → company.x_fc_default_tz → UTC) and the formatters use
Intl.DateTimeFormat with the explicit timeZone option. plant_overview
now consumes server_time (already fp_format'd) instead of toLocaleTime
String. Same chain Odoo backend uses, so PDF / view / tablet all
agree on what time it is.
Versions: fusion_plating_jobs 19.0.10.30.0,
fusion_plating_shopfloor 19.0.33.1.12.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Per user request: technicians on the tablet should only see Discuss,
To-do, Plating, AI, Maintenance, Time Off. Every other top-level app
menu (Calendar, Contacts, CRM, Sales, Dashboards, RC, Faxes, Field
Service, Fusion Clock, Invoicing, Accounting, Project, Timesheets,
Planning, Shipping, Website, Purchase, Inventory, Sign, HR, Payroll,
Attendances, Recruitment, Expenses, IoT, Link Tracker, Apps) is now
restricted to a new group_fp_office_user.
Architecture:
- New group_fp_office_user (security/fp_menu_visibility.xml) — a
marker group that controls back-office menu visibility.
- Owner / Manager / Quality Manager / Shop Manager / Sales Rep all
imply office_user via implied_ids — they see everything they did
before.
- Pure Technicians do NOT imply office_user — they see only the
tablet-friendly menus.
- A "!technician" filter would have hit managers too (because Manager
→ ... → Technician via implication), so office_user is the inverse
pattern that gets the right scoping.
Implementation:
- post_init_hook + migrations/19.0.21.4.0/post-migrate.py both call
_fp_apply_office_user_menu_visibility(env) which iterates a curated
list of menu xmlids and sets group_ids = [office_user] on each.
- Uses env.ref(..., raise_if_not_found=False) so menus from
uninstalled modules silently skip — no hard depends added.
- ir.ui.menu uses `group_ids` in Odoo 19 (was groups_id pre-18 — same
rename pattern as res.users; CLAUDE.md Rule 13c).
- Settings / Apps / Tests left untouched (already admin-restricted).
- Some menus (Field Service) end up with office_user OR their original
group — that's correct behavior: Plating Techs have neither so still
don't see them; explicit Field Technicians keep access.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Adds [-] / [+] buttons around every numeric input in the Record Inputs
dialog (single-value, dual-entry, and pass_fail+range branches). Tap
to increment / decrement by the recipe-author-derived step size
(stepFor() already computes this from target_min/target_max precision,
falling back to input-type defaults).
- Decrement clamps at 0 (typical qty/time/temp on a plating floor
doesn't go negative; if needed, operator can still tap the input
and type a negative value)
- Increment uses _stepRound() to avoid floating-point fuzz on decimals
- Center-aligned monospace-ish input between the buttons for clarity
- inputmode='decimal' (or 'numeric' for time fields) hint so when the
operator does tap the input, the iPad shows a number keypad instead
of the full keyboard
Touches single-value, dual-entry (min/max), and pass_fail+range. Other
multi-field widgets (multi-point thickness, bath chemistry panel) still
use plain inputs — separate request if they need steppers too.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Native browser confirm popups look out of place in the tablet UI.
Mark Counted is already a deliberate prior step, so requiring a
second confirmation on Close Receiving was just friction. If a
receiver hits Close prematurely, action_reset_to_counted on
fp.receiving from the back office is the recovery path.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Two bugs:
1. Gate silently passed when step.recipe_node_id was NULL — happened
to every WO-30057 step after this morning's clone delete (the FK
ON DELETE SET NULL wiped the link). _fp_missing_required_step_inputs
returned an empty recordset when node was None, so the gate had
nothing to fail on and button_finish succeeded with zero audit.
Fix: _fp_check_step_inputs_complete now treats NULL recipe_node_id
as an explicit "no recipe link" hard block. Operator can't finish;
manager bypass posts chatter audit.
2. No tablet UI for the manager bypass. The gate's bypass was a
Python context flag — invisible from the JS layer, so managers
were stuck behind the same hard error as operators.
Fix: new /fp/workspace/finish_step endpoint returns structured
errors (gate type, missing_prompts list, bypass_available bool).
Server-side enforces manager group when bypass=True (can't trust
the client). New FpFinishBlockDialog OWL modal renders:
- Non-manager: Cancel + Record Inputs
- Manager: Cancel + Record Inputs + ⚠ Bypass & Finish (audit)
JobWorkspace.onFinishStep routes plain finishes through the new
endpoint; signature-required steps still go through /fp/workspace/sign_off
(separate gate). Added is_manager to /fp/workspace/load payload so
the JS knows which dialog variant to render.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Pure client-side tick — 1s setInterval bumps state.tickNow which the
template reads via formatActiveStepElapsed(step). No RPC per tick.
Reads step.date_started_iso (UTC) from the existing payload, parses
to ms, displays elapsed since.
- Green pill (#d1fae5 bg, monospace tabular-nums) on the ACTIVE badge
- Flips red (#fee2e2 + pulse animation) when elapsed > 1.5x
duration_expected — visual cue for the operator that the step is
running long against the recipe target
Cleanup interval on onWillUnmount alongside the existing 15s refresh
interval.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Adds the receiver workflow to the Job Workspace tablet view (was the
gap behind WO-30057 sitting in Receiving with no way to advance).
Receivers no longer need to go to the backend form.
Workspace card (renders above the step list when fp.receiving in
state draft/counted on the linked SO):
- Draft state: numeric box-count input + per-line received_qty /
condition picker (good/damaged/mixed) + Damage Log panel + Mark
Counted button. Autosaves on input blur.
- Counted state: read-only summary (boxes, parts, who/when) +
Damage Log still editable + Close Receiving button.
- Closed: card disappears, recipe takes over.
New FpDamageDialog OWL modal:
- Severity pill picker (Cosmetic / Functional / Rejected) with
color-coded active state
- Required description textarea
- Action Required pill picker (None / Notify / Return / As-Is)
- Photo capture: both "Take Photo" (input capture="environment"
triggers tablet camera) AND "Upload" (file picker fallback).
Multi-photo with preview grid + per-photo remove.
5 new endpoints on workspace_controller.py:
- receiving_save_lines (autosave box_count_in + per-line qty/cond)
- receiving_mark_counted (wraps action_mark_counted)
- receiving_close (wraps action_close)
- damage_create (creates fp.receiving.damage + attaches base64 photos)
- damage_delete (removes a damage row)
No model changes — wraps existing fp.receiving actions and damage
CRUD. C3 (outbound shipping carrier/label) is a separate spec.
Spec: in-conversation brainstorm (C1+C2) following the 2026-05-24
workspace step actions spec; no standalone doc since scope is small.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Fix: in the Job Workspace tablet view, the Start button was buried
inside a parent t-if that required the step to already be in_progress
or blocked. So ready/paused steps showed no buttons at all -
operators couldn't advance the WO from this screen (the reason the
user couldn't complete anything on WO-30057).
Template restructure (job_workspace.xml):
- Always-visible line 1 (icon + step# + name + ACTIVE/PAUSED badge + meta)
- Non-terminal detail panel (chips + instructions + opt-out + GateViz)
visible on every non-done step so operator reads ahead
- Action row dispatched per-kind via getStepActions() helper
Per-kind action dispatcher (job_workspace.js):
- in_progress -> Record Inputs, Pause, Finish (or Finish & Sign Off)
- paused -> Resume, Record Inputs, Finish
- contract_review (ready) -> Open QA-005 Form
- gating (ready) -> Mark Passed (1-click start+finish)
- requires_rack_assignment -> Start (Assign Rack) - opens FpRackPartsDialog
- else (ready) -> Start
5 new handlers: onPauseStep / onResumeStep / onMarkPassed /
onOpenContractReview / onStartWithRack. Pause and Resume use ORM RPC
(button_pause/button_resume) since no HTTP endpoint exists.
New model method (fp.job.step.action_mark_gating_passed):
- 1-click pass for gating steps - does button_start + button_finish
in one transaction, posts chatter "Gate X marked passed by Y"
- Raises UserError if called on a non-gating step (defensive)
- Bypasses S21 required-inputs gate (gating steps have no inputs)
Controller: workspace_controller.py adds requires_rack_assignment to
the step payload so the JS dispatcher can route correctly.
Spec: docs/superpowers/specs/2026-05-24-workspace-step-actions-design.md
Sub-B (Record Inputs tablet polish: inputmode/prefill/date pickers/
signature pad/camera) is brainstormed but deferred.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Root causes documented:
1. Recipe 3620 ENP-ALUM-BASIC had duplicate sequences (Contract
Review + Masking both at seq 10; Incoming Inspection + Racking
both at seq 20). Clones inherited the ambiguity and resolved by
id ordering, putting Masking before Incoming Inspection — which
meant new jobs went straight to Plating column after the
contract-review auto-complete.
2. 24 per-part clone recipes accumulated, all carrying the broken
ordering.
3. ~10 kind=other stragglers across the base recipes (Blasting,
Adhesion Test Coupon, Strip Process, Chemical Conversion etc.)
4. Recipe duplication had no kind safety net.
Implementation shipped in commits referenced from the plan.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Two follow-up fixes caught during the entech deploy of recipe cleanup:
1. RESOLVER_KIND_TO_ACTIVE_KIND was missing a self-pass entry for
'wet_process'. The new aliases added in 19.0.21.3.0 (Chemical
Conversion, Trivalent Chromate Conversion, Strip Process - AL,
Plug The Threaded Holes via mask) directly return 'wet_process'
from the resolver — without the passthrough they didn't translate
to any active kind and stayed as 'other'. Added 'wet_process':
'wet_process' so the migration's Phase 2 backfill catches them.
2. Migration 19.0.10.26.0 Phase 3 was using batch unlink
(clone_recipes.unlink()) which tripped a PostgreSQL FK cascade
ordering bug on entech ("insert or update on parent_id violates
FK ..." during the CASCADE chain). Rewrote Phase 3 to delete one
clone at a time with SAVEPOINT per clone — slower but immune to
the batching bug, and one failed clone doesn't roll back the
whole transaction.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Migration 19.0.10.26.0/post-migrate.py runs in 5 phases:
1. Resequence recipe 3620 ENP-ALUM-BASIC ops to fix the duplicate-
sequence bug (Contract Review=10, Incoming Inspection=20,
Masking=30, Racking=40, then the rest). Also delete the empty
duplicate ENP-Alum Line sub_process (id 4056).
2. Backfill kind on all kind=other nodes via the extended resolver
from fusion_plating 19.0.21.3.0
3. Delete all per-part clone recipes (name contains em-dash)
4. Recompute fp.job.step.area_kind on all steps
5. Recompute fp.job.active_step_id + card_state on in-flight jobs
Plant kanban: no_parts cards now always land in the Receiving column
regardless of active_step area_kind. The receiver works Receiving;
that's where the card belongs when parts haven't arrived.
Spec: docs/superpowers/specs/2026-05-24-recipe-cleanup-design.md
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Resolver (fp_resolve_step_kind) extensions:
- New aliases: blasting/bead blast/media blast variants, adhesion
testing, corrosion testing, lab testing, strip process, chemical
conversion, trivalent chromate, plug the threaded holes, air dry,
desmut, soak clean, cleaner, nickel strike/strip
- Parenthetical suffix stripping - "Masking (If Required)" resolves
through "masking", "Incoming Inspection (Standard)" through
"incoming inspection", "Trivalent Chromate Conversion (A-14 / A)"
through "trivalent chromate conversion"
- New RESOLVER_KIND_TO_ACTIVE_KIND map translates the resolver's
vocabulary (cleaning/electroclean/etch/rinse/strike/dry/wbf_test
-> wet_process) so the resolver output lands on active kinds only
Auto-classify hook on fusion.plating.process.node:
- _fp_autoclassify_kind() upgrades kind_id when current is 'other'
AND name resolves via the resolver. Idempotent - never overrides
a non-'other' kind. Skip via context flag fp_skip_kind_autoclassify
- Wired into create() and write() (only fires when name or kind_id
changed on write)
- Side-effects: recipe duplication via copy() auto-corrects newly
copied nodes; Simple/Tree editor authoring auto-classifies as soon
as the name is saved
Spec: docs/superpowers/specs/2026-05-24-recipe-cleanup-design.md
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Required because fp.job.card_state (stored) has @api.depends including
active_step_id.area_kind. When step.area_kind changes, Odoo's trigger
chain searches fp.job by active_step_id — non-stored fields can't be
queried in WHERE clauses, raising ValueError("Cannot convert ... to
SQL because it is not stored").
Caught during entech deploy of 19.0.10.25.0/post-migrate.py Phase 3
(steps._compute_area_kind() failed on first run). store=True makes
the column searchable and the trigger chain works.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
- fp.step.kind.area_kind: drop tracking=True (model doesn't inherit
mail.thread; tracking was a no-op emitting a startup warning).
- Migration 19.0.10.25.0: anchor the De-Masking ILIKE so it doesn't
wildcard-match "Ready For De-Masking" (which the earlier "Ready %"
rule already routes to gating). Also drop the cur_code='mask' filter
so the 4 De-Masking nodes still classified as 'other' get picked up
on fresh re-runs too.
Direct SQL applied the 4-row fix on entech (post-migrate doesn't
re-run for already-applied versions); this commit keeps fresh
installs and any future re-runs consistent.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Fix the Shop Floor plant kanban so cards land in the right column:
- fp.job._compute_active_step_id walks priority chain
(in_progress > paused > ready > pending), not just in_progress
- fp.job._compute_card_state edge case respects job.state='done'
(no more bogus 'contract_review' label on done jobs)
- fp.job.step._compute_area_kind reads kind.area_kind directly;
legacy _STEP_KIND_TO_AREA dict removed (50+ lines deleted)
- /fp/landing/plant_kanban filters out done/cancelled jobs from
the live board
Migration 19.0.10.25.0 backfills template metadata (codes,
descriptions, icons, kind_id) on 30 unfinished library templates
and repoints recipe nodes for 6 unambiguous name patterns
(Blasting -> blast, Ready For X -> gating, De-Masking -> demask,
Scheduling -> gating, Nickel Strip -> wet_process,
Pre-Meas/Check Sulfamate -> inspect).
Battle test bt_s24_between_steps.py covers between-step routing,
paused step lifecycle, and done-job board filter.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Add required area_kind Selection to fp.step.kind so each kind
self-declares which plant-view column its steps belong in. Replaces
the hardcoded _STEP_KIND_TO_AREA dict (removed in fp_job_step.py
in the follow-up commit).
- New `blast` kind for the Blasting column (sequence=35)
- 26 existing kind records seeded with area_kind in XML
- Pre-migrate 19.0.21.2.0 seeds existing rows BEFORE NOT NULL hits
the schema; also activates derack/demask/gating that were
deactivated in 19.0.20.6.0 but are needed for the full taxonomy
- Step Kind form + list views surface area_kind (badge + chip)
- Step Kind search adds Group By Shop Floor Column
- Simple Editor kind picker shows "Masking — Masking column"
suffix so authors see the routing at pick time
- Add Hot Water Porosity Test (A-15) + Final Inspection / Packaging
templates (used by 7+3 recipe nodes that previously had no
library entry)
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Odoo 19's Session.authenticate(env, credential) takes an Environment as
the first arg, not a db-name string. Passing request.db triggered
TypeError: 'str' object is not callable on the internal
env(user=None, su=False) reset.
Fixes the "Odoo Server Error" dialog operators saw when trying to PIN
unlock from the tablet. Same fix applies to lock_session (which was
silently masked by its broad except Exception).
Bumps fusion_plating_shopfloor to 19.0.33.1.2.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Pure code change (no DB schema), but bumping the patch version
keeps repo manifest aligned with the deployed state so the next
-u doesn't no-op due to version match.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The kiosk_login in /fp/tablet/lock_session was hardcoded to the
data XML's original value ('fp_tablet_kiosk@enplating.local'). The
data record is noupdate='1', so admins can (and on entech, did)
rename the kiosk user on the form for memorability — the rename
persists through -u, but the hardcoded string in the controller
silently breaks the re-auth-as-kiosk path.
Fix: resolve the kiosk login dynamically via env.ref of the xmlid
'fusion_plating_shopfloor.user_fp_tablet_kiosk'. Robust against any
future rename. CLAUDE.md updated to make 'identify by xmlid, never
by login string' an explicit convention for the tablet flow.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Tablet PIN session redesign Phase G removed all tablet_tech_id
plumbing. CLAUDE.md still documented the old session-pool + kwarg
flow which would mislead future-Claude. Updated to describe the
new per-tech-session attribution + kiosk re-auth flow, plus the
gotcha about keeping ir.config_parameter['fp.tablet.kiosk_password']
in sync with the actual user-record password.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Records the legacy-tablet-flow-removed state. Triggers -u so the
module's installed version reflects the post-cleanup code (the
ir_module_module row shows 19.0.33.1.0 after deploy).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Frontend cleanup completing Phase G of the tablet PIN session
redesign:
- tablet_lock.js: removed sessionMode branching (no legacy path).
unlock() always calls /fp/tablet/unlock_session + reloads.
handOff() always calls tabletSessionManager.lockBack('manual').
isLocked uses currentUid vs kioskUid exclusively. _checkIdle
still drives the warning UI via activity_tracker; the actual
lock RPC is owned by tablet_session_manager.
- fp_rpc.js: simplified to a thin async pass-through around @web/core
network rpc. tech_store-based tablet_tech_id injection is gone
(the session uid IS the tech).
- tech_store.js: DELETED (replaced by per-session backend attribution
+ tablet_session_manager for lock state). Removed from manifest.
- Wrapper components (shopfloor_landing, job_workspace,
manager_dashboard, plant_kanban): swapped useService('fp_shopfloor_tech_store')
for useService('fp_tablet_session_manager'); techStore.lock() ->
tabletSessionManager.lockBack('manual'). plant_kanban's defensive
try/catch on the tech_store lookup is no longer needed.
- tablet_lock.xml: Hand-Off button no longer gated on sessionMode;
always rendered.
- Tests: removed legacy TestTabletUnlock class from test_tablet_pin.py
(covered the deleted /fp/tablet/unlock route). Dropped session_mode
assertion from test_tiles_bootstrap_fields.py (the return key is
gone post-Phase-G). kiosk_uid + current_uid assertions retained.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Session-swap is now the only flow. Legacy /fp/tablet/unlock endpoint
deleted. _tablet_audit.py (env_for_tablet_tech helper) deleted with
its last caller gone. /fp/tablet/ping no longer takes current_tech_id
(session uid IS the tech). /fp/tablet/tiles drops tablet_session_mode
return key (kiosk_uid + current_uid retained for OWL isLocked logic).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Session swap makes attribution automatic via request.env.user — the
tablet_tech_id plumbing is dead code after the kiosk + per-tech-session
architecture lands. Removed kwarg from 3 endpoints in
manager_controller, 11 in shopfloor_controller, 3 in
workspace_controller. _tablet_audit.env_for_tablet_tech import gone
from all 3 files.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Owner-only smart button on res.users form. Click opens the audit log
filtered to that user (both user_id and attempted_user_id, so
failed unlock attempts against a tile show up too).
Compute is non-stored: search_count on the audit model per user on
demand. Sudo'd because the audit model has Owner-only ACL — the
compute fires for the form-viewing user (Owner) who would see the
results anyway via the menu.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Plating > Configuration > Tablet Audit Log. Read-only list with
decoration (green=unlock, red=failed, warning=ceiling/force,
muted=manual/idle). Form shows full forensic detail incl. ip/ua.
Owner-only via groups=fusion_plating.group_fp_owner on the menu.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Important I1: tablet_session_manager.beginSession() now calls
_removeListeners() (and clears any pending _tickHandle) defensively
at start. Prevents DOM listener leak on dev hot-reload or any path
that re-bootstraps without a clean endSession() first.
Important I2: tablet_lock._checkIdle() early-returns in session_swap
mode. The tablet_session_manager owns idle tracking there (5s poll,
calls /fp/tablet/lock_session directly). Was previously dormant by
accident because session_swap never populates the legacy techStore;
explicit guard makes the decoupling intentional.
Minor M5: session_swap unlock success now resets selectedTileUserId
before window.location.reload(), matching the legacy path''s
cleanup pattern. Cosmetic before reload kicks in.
Minor M9: New test_tiles_bootstrap_fields with 3 HttpCase tests
asserting /fp/tablet/tiles returns tablet_session_mode, kiosk_uid,
and current_uid. The OWL lock screen branches on all three — a
contract regression would silently break session_swap.
Minor M10: Added inline comment near _sessionModeCache declaration
in fp_rpc.js explaining the page-reload-invalidates-cache lifecycle.
Deferred (for future polish, NOT in this commit):
- I3 (_getSessionMode ACL gap for tech users — functionally correct,
just suboptimal; cache fallback to ''legacy'' kicks in)
- M6 (wrapper component Hand-Off buttons no-op in session_swap)
- M7 (hardcoded idle/ceiling thresholds — server-configurable later)
- M8 (timer divergence vs activity_tracker — unify later)
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
When tablet_session_mode='session_swap', the server attributes every
write via request.env.user — there's no need to pass tablet_tech_id
in the RPC params. Caches the mode lookup at module level so we don't
round-trip on every RPC.
Legacy mode unchanged — fpRpc still injects tablet_tech_id from
techStore.currentTechId.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
When ir.config_parameter[fp.shopfloor.tablet_session_mode]='session_swap',
PIN submit calls /fp/tablet/unlock_session and reloads the page; the
new session manager service kicks in on next mount. handOff() calls
lockBack('manual') which destroys the tech session server-side and
re-auths as kiosk.
Legacy mode unchanged — same /fp/tablet/unlock + techStore flow.
The feature flag, kiosk_uid, and current_uid arrive via the existing
/fp/tablet/tiles bootstrap response (Task D0).
Adds a tablet_lock-owned Hand-Off button visible only in session_swap
mode (in legacy mode wrapper components own their own buttons that hit
techStore.lock(); session_swap renders our own button so the manual
hand-off goes through lockBack() + page reload).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Tracks idle + ceiling timers for an unlocked tech session. Fires
/fp/tablet/lock_session when either trips, then reloads the page so
the browser re-bootstraps under the fresh kiosk session.
Defaults: 10min idle, 8hr ceiling, 5s tick interval. Listens for
click/touchstart/keydown/mousemove as activity signals.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
OWL lock screen needs to know (a) the active session mode (legacy or
session_swap) to branch between endpoints, and (b) the kiosk uid to
determine 'is the current browser session the kiosk?' Both come from
the bootstrap response so no extra round-trips on every render.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Important 1: lock_session now closes the original unlock event's
session_ended_at via the same parameterized-SQL bypass pattern used
by the force-lock cron. Without this, every Hand-Off click became
a duplicate force_lock event 8 hours later (cron saw the unlock still
open and re-processed).
Important 2: test_unlock_lock_session_endpoints setUp now
unconditionally overrides the kiosk password (was gated on
'if not get_param(...)' which broke on entech where the post-migrate
hook already generated a random password — tests failed against the
real value). HttpCase rolls back per test so no persistence.
Minor 4: _cron_force_lock_stale_sessions now routes the force_lock
create through write_event helper for consistency (single audit-write
path; helper captures acting_uid/ip/ua uniformly).
Minor 5: Hoisted local imports inside method bodies to top-of-file
in tablet_controller.py (AccessDenied, _tablet_session_audit) and
fp_tablet_session_event.py (timedelta, write_event).
Minor 6: New test_force_lock_cron.py with 3 tests: stale session
emits force_lock + closes original; recent session unaffected;
already-closed session not re-processed. Would have caught
Important 1 if it had existed during Phase C review.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Every 5 minutes, find active unlock events past 8-hour ceiling and
mark them force-locked. SQL bypass of the model's read-only ACL is
the only path that can update existing rows (no Python write() works
because the model override blocks even sudo writes without the
explicit fp_tablet_audit_admin_write context flag).
Ceiling configurable via ir.config_parameter[fp.tablet.session_ceiling_hours].
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Writes lock event (manual/idle/ceiling) with duration computed from
the open unlock event. Then logout + re-authenticate as kiosk via
the password stored in ir.config_parameter['fp.tablet.kiosk_password'].
Falls back to 'needs_kiosk_relogin' if the kiosk password is missing
(sysadmin must log in manually). Logs every event for forensic
review.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
PIN verify -> request.session.authenticate(type=fp_tablet_pin) -> new
session sid, cookie swap, audit event written. Failed attempts also
written to audit log (failed_unlock, failure_reason=wrong_pin or
locked_out or no_pin_set or user_inactive).
OLD /fp/tablet/unlock stays alive during the 1-week overlap window
per spec Section 5.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Single source for sha256(session sid), ua trim, ip/acting_uid capture
from request. Used by unlock_session, lock_session, and force-lock cron.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
C1: Add placeholder fp_tablet_cron.xml + fp_tablet_session_event_views.xml
so the module is installable now (real content lands in Phase C task C4
and Phase E task E1 respectively).
I1: test_tablet_pin_auth_manager now passes {} (not self.env) as the
env arg to _check_credentials — matches what request.session.authenticate
provides and what the base implementation expects.
I2: Auth manager role check now uses user_sudo.all_group_ids (transitive)
instead of group_ids (direct) per CLAUDE.md rules 13l + 23. Owner users
who hold Owner directly still match all 5 shop-branch xmlids via the
implication chain.
I3: fp.tablet.session.event gains Python-layer write() + unlink()
overrides that always raise AccessError unless the explicit
fp_tablet_audit_admin_write / fp_tablet_audit_admin_purge context flag
is set. Closes the gap between the model's append-only docstring and
its actual enforcement (ACL-only previously).
M1: Hoisted 'from odoo.exceptions import AccessDenied' to top-of-file
imports next to existing UserError import.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
8 tests: correct/wrong/missing PIN, missing/unknown login, inactive
user, no shop-branch role, and pass-through of other credential types.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Validates PIN hash + shop-branch role membership when the credential
type is fp_tablet_pin. Goes through Odoo's standard _check_credentials
chain so future 2FA / IP-gate modules layer cleanly on top.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Code-review findings on Phase A (Tablet PIN Session Redesign):
I1: Security XML comment now honestly describes the kiosk as Internal
User + explicit reads, not 'near-zero ACL'. base.group_user is kept
(required for auth='user' HTTP routes to function) but the comment
no longer overstates how locked-down the kiosk is.
I2: New ir.rule scopes the kiosk's ir.config_parameter read to keys
matching 'fp.tablet.%' or 'fp.shopfloor.%'. Combined with the
existing model-level read ACL, kiosk can no longer enumerate
third-party secrets (e.g. fusion_tasks.vapid_private_key) or
arbitrary API keys stored in ICP.
I3: post-migrate docstring now advises sysadmins to unlink the
plaintext ICP password row after kiosk tablets are paired, to
minimise plaintext-in-backups risk. Rotation procedure documented.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Generates a random kiosk password on first deploy, stores in
ir.config_parameter for sysadmin retrieval. Idempotent — re-runs
on subsequent -u leave the password alone.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Kiosk holds the tablet session when no tech is PIN-unlocked.
Password is auto-generated by the post-migrate hook (Task A5).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
7 phases (A-G), ~25 tasks. Phase A-E build the new auth flow,
audit model, endpoints, OWL service, and audit UI. Phase F is the
entech rollout (manual, inline by main session per hybrid pattern).
Phase G is the post-overlap cleanup (rip out tablet_tech_id,
delete legacy endpoint, archive shopfloor service user).
Bakes in 7 known gotchas from the permissions overhaul (rules
13c, 13i, 13k, 13m, 13d, AUDIT-1, always-push-to-main) so the
implementer doesn't repeat them.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Real per-tech Odoo sessions on PIN unlock (not just attribution).
Closes the audit-trail gap from Phase 1 permissions overhaul: today
the tablet runs as a persistent 'shopfloor service' user and the PIN
is just an OWL overlay — every action is attributed to whoever the
session user is, not the tech who tapped their tile.
Locked decisions:
1. Real per-tech sessions (impersonation, cookie swap)
2. Idle timeout 10min + manual lock + 8hr hard ceiling
3. Dedicated kiosk user (fp_tablet_kiosk, near-zero ACL)
4. No manager override — Mgr/Owner PIN in as themselves
5. Two-step deploy with 1-week overlap; OLD endpoint removed after
successful rollout
Audit: fp.tablet.session.event append-only log captures unlock /
manual_lock / idle_lock / ceiling_lock / force_lock / failed_unlock
/ admin_reset events with ip, ua, session hash, duration.
Effort: ~4 dev days + 1 week observation. Plan via writing-plans
skill next.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Same mistake as the original implementer wave — used the deprecated
groups_id field name on res.users in the search domain. Odoo 19 raises
ValueError: Invalid field res.users.groups_id. Should be group_ids.
CLAUDE.md rule 13l example also fixed so future-Claude doesn't copy
the bug from the documentation.
Module version: 19.0.32.0.12 -> 19.0.32.0.13
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Previously only direct Technicians appeared on the lock-screen tile
grid because env.ref('group_fp_technician').user_ids returns DIRECT
members only — Odoo's implication chain (Owner -> ... -> Technician)
is read-time only, not stored in res_groups_users_rel.
Search res.users with ('groups_id', 'in', shop_branch_ids) where
shop_branch_ids covers all 5 shop-branch role groups (Technician,
Shop Manager v2, Manager, Quality Manager, Owner). Sales branch
intentionally excluded — they don't operate the tablet.
Verified on entech: 18 technicians + 1 shop_manager + 2 managers
+ 1 quality_manager + 2 owners = 24 tiles (was 18).
CLAUDE.md rule 13l corrected — previous version wrongly claimed
res.groups.user_ids surfaced implied members. Now documents the
search-based query as the canonical 'enumerate role X or higher'
pattern.
Module version: 19.0.32.0.11 -> 19.0.32.0.12
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Operators trying to Finish a step with required step_input prompts
got the S21 gate error telling them to 'Click Record Inputs on the
step row' — but the workspace UI never exposed that button. Only the
job-form view had it.
Adds a 'Record Inputs' secondary button next to Finish/Finish & Sign
Off when the step is active. Click opens the fp_record_inputs_dialog
(via action_open_input_wizard on fp.job.step). On dialog close the
workspace refreshes so the step's progress chip updates.
Module version: 19.0.32.0.10 -> 19.0.32.0.11
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Same pattern as plant_kanban — workspace payload denormalizes
cross-module fields Technician can't read directly (sale.order,
fp.part.catalog, customer_spec, etc.). job.sudo() at the top so
the whole render path is sudo'd.
Job Workspace was stuck on 'Loading...' with a server-error toast
because the route returned {ok:false, error:'...'} (27-byte response)
when the first cross-module field access AccessError'd.
Module version: 19.0.32.0.9 -> 19.0.32.0.10
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Post-migration, Technicians (now group_fp_technician) have read on
fp.job but NOT on sale.order / fp.part.catalog / fusion.plating.customer.spec.
The kanban render path tries to access job.sale_order_id.x_fc_po_number
and AccessErrors silently — kanban returns empty, user sees blank
'Shop Floor' page.
Fix: `job = job.sudo()` at the top of _render_card. The output is
denormalized display data, no security concerns; ACL gating is still
enforced by the caller's access to fp.job (which Technician does have).
CLAUDE.md rule 13m documents the broader pattern: any dashboard /
tablet / kanban controller surfacing cross-module data to low-priv
roles needs this sudo at the helper top.
Module version: 19.0.32.0.8 -> 19.0.32.0.9
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Operators read phone-style clocks; 24-hour was off-norm for North
American shop. Hour no longer zero-padded (1:05 PM, not 01:05 PM)
to match the iPhone/Android idiom.
Module version: 19.0.32.0.7 -> 19.0.32.0.8
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Wider tablets fit 5 tiles per row comfortably; 3 was too sparse with
a 20-person operator roster (forced a long vertical scroll). Bumped
.o_fp_lock_tiles max-width from 480px to 800px so the tiles don't
stretch wide at 5 columns.
Module version: 19.0.32.0.6 -> 19.0.32.0.7
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
res.config.settings.x_fc_default_landing_action_id is related= to
res.company.x_fc_default_landing_action_id, which was widened from
ir.actions.act_window to ir.actions.actions in the Phase I post-deploy
fixes (so the picker accepts both window AND client actions). The
settings field's comodel was left at the old type and tripped on
opening Settings: 'Wrong value for ...: ir.actions.actions()' when
the related compute tried to write the client-action value into the
narrower settings field.
Module version: 19.0.21.1.2 -> 19.0.21.1.3
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Consolidates 12 res.groups into 8 clean roles:
Owner -> Quality Manager -> Manager -> [Shop Manager, Sales Manager]
-> [Technician, Sales Rep], plus implicit 'No' (no plating group).
Phase A — 7 new res.groups with implied_ids chains + backward-compat;
old groups marked [DEPRECATED] and queued for 30-day cron purge.
Phase B — mechanical ACL sweep across 24 ir.model.access.csv files.
Phase C — Manager/QM quality permission split + FAIR/Nadcap ir.rule.
Phase D — 3-layer menu/submenu/field/button visibility hardening.
Phase E — role-based landing-page dispatch (Owner -> Manager Desk,
QM -> Quality Dashboard, Sales Rep -> Quotations, Tech -> Plant
Kanban, etc.) + picker domain over ir.actions.actions so window AND
client actions are both pickable.
Phase F — Owner-only Plating > Configuration > Team kanban for
drag-and-drop role assignment, plus Designated Officials (CGP DO +
Nadcap Authority) fields on res.company.
Phase G — Sales Manager + required to confirm SO; fixed the
audit-finding-11 _administrator typo that had made the account-hold
bypass dead code; swept all Python has_group() refs to new xmlids.
Phase H — dry-run + Owner-approval migration workflow with
fp.migration.preview model, mail.activity notification, 30-day
rollback window, daily purge cron.
Phase 9 — final-reviewer fixes (groups_id->group_ids, server-action
wiring, migrations/19.0.21.1.0/post-migrate.py for -u dispatch,
Odoo 19 kanban card template, FAIR/Nadcap cert_type field name,
user_has_groups removed from invisible attrs).
Phase I — pre-deploy backup, entech deploy (5 cascade fixes
discovered live), Owner approval of migration #1 (25 users
migrated cleanly), post-approval SQL verification, sample login
tests, deprecated-group picker cleanup (Option A SQL UPDATE),
and 11 post-deploy bug fixes (picker model swap to ir.actions.actions,
ACL grant for ir.actions.actions read to plating users, SELF_WRITEABLE_FIELDS
extension for non-admin Preferences save, res.users.message_post ->
partner_id.message_post, tablet lock screen group ref swap,
PIN-pad dark-mode dot contrast, lock-screen logo plate dark mode).
Spec: docs/superpowers/specs/2026-05-23-permissions-overhaul-design.md
Plan: docs/superpowers/plans/2026-05-24-permissions-overhaul-phase1-plan.md
CLAUDE.md rules added: 13b-13l (Odoo 19 gotchas surfaced during build/deploy)
Live state on entech: 25 users migrated, 30-day rollback open
until 2026-06-23, deprecated groups hidden from picker.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
8 distinct bugs caught + fixed while testing the live admin DB on entech
after the migration was approved. Each surfaced a real Odoo 19 gotcha
now codified in CLAUDE.md (rules 13b-13l).
Picker architecture:
- res.users.x_fc_plating_landing_action_id and res.company.x_fc_default_landing_action_id
now Many2one('ir.actions.actions') instead of ('ir.actions.act_window'),
so the picker accepts BOTH window actions (Sale Orders / Quotations /
Process Recipes) AND client actions (Manager Desk / Plant Kanban /
Quality Dashboard). Picker went from 3 entries to 6.
- x_fc_pickable_landing field moved from the two subclasses to the
ir.actions.actions base. Single source of truth.
- _render_resolved on the base dispatches to the correct subclass by
action type.
Non-admin Preferences access:
- Added ACL grant: group_fp_technician (and all higher roles via
implication) get read on ir.actions.actions. Without this, opening
Preferences raised AccessError on the picker domain evaluation.
- Removed the accessible_landing_action_ids Many2many compute (failed
for non-admins because field assignment requires write access on
the comodel relation, even with sudo'd search). Picker now shows all
6 entries to all users; resolver falls through gracefully if the
user picks an action they can't reach.
- res.users SELF_WRITEABLE_FIELDS / SELF_READABLE_FIELDS extended via
@property + super() (NOT class attribute — Odoo 19 changed the
pattern). Non-admin users can now save the Preferences dialog with
plating fields without hitting the standard write ACL.
Migration workflow:
- res.groups.users -> .user_ids (Odoo 19 rename; deprecated alias
removed). Was crashing _fp_notify_owners and _cron_purge_expired.
- user.message_post -> user.partner_id.message_post (res.users uses
_inherits delegation which doesn't expose mail.thread methods).
Was crashing the Owner approval click.
Tablet lock screen:
- /fp/tablet/tiles points at group_fp_technician instead of the old
group_fusion_plating_operator. Post-migration nobody holds the old
group directly (only via implication), so res.groups.user_ids on
the old xmlid returned empty — 'No operators configured' shown
even with PIN set.
- PIN pad dots dark mode: empty dot now dark gray (#424245), filled
dot now pure white. Previous version had both at light shades so
user couldn't see PIN entry progress.
- Lock-screen logo frame dark mode: near-opaque white plate
(rgba 0.95) so company logos designed for light backgrounds
render correctly. Previous 0.08 alpha let the dark page bleed
through.
Pre-deploy collision fix (already committed before deploy but
documented here for completeness):
- pre-migrate.py to rename old configurator's 'Shop Manager' group
display name before new fp_security_v2.xml loads the new
group_fp_shop_manager_v2 with the same display name (avoids
res_groups_name_uniq violation).
Module versions bumped:
fusion_plating: 19.0.21.1.0 -> 19.0.21.1.2
fusion_plating_shopfloor: 19.0.32.0.4 -> 19.0.32.0.6
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
5 fixes discovered during the live deploy to entech LXC 111:
1. pre-migrate.py to rename old configurator's 'Shop Manager' group BEFORE
new core 'Shop Manager v2' XML loads (cross-module name collision on
res_groups_name_uniq).
2. res_company_views.xml: dropped ref() inside <field domain=> attribute
(Odoo 19 view validator interprets it as a field name).
3. sale_order_views.xml: replaced 3 separate xpaths for amount_total /
amount_untaxed / amount_tax with a single xpath on tax_totals widget
(Odoo 19 sale.view_order_form uses one widget instead of separate fields).
4. fp_cert_security.xml: certificate_type field, not cert_type. FAIR is a
separate model so the rule only restricts cert_type='nadcap_cert' now.
5. fp_certificate_views.xml + fp_capa_views.xml + fp_customer_spec_views.xml:
stripped user_has_groups() from invisible= / readonly= attrs (Odoo 19
view validator interprets as field name). Model-layer ACLs and ir.rules
already enforce the same restrictions.
Also fixed res.groups.users -> user_ids in fp_migration.py (Odoo 19 rename,
caught when manually invoking _fp_notify_owners post-deploy).
CLAUDE.md updated with 4 new rules (13e cross-module name collisions,
13f ref() in domain, 13g tax_totals widget, 13h user_has_groups in attrs).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Pre-deploy fixes for Phase 1 permissions overhaul branch (catches by
final-reviewer subagent + main session).
CRITICAL FIXES:
C1: groups_id -> group_ids (Odoo 19 field rename). Affected ~30 sites
across 4 model files, 1 view, 7 test files. Documented project
gotcha (feedback_odoo19_groups_id_renamed.md) that the implementer
subagents missed because they don't see user memory.
C2: action_fp_resolve_plating_landing server action now calls
env['ir.actions.act_window'].sudo()._fp_resolve_landing_for_current_user()
instead of the old inline priority chain. Phase E's role-based
dispatch was previously dead code.
C3: New migrations/19.0.21.1.0/post-migrate.py triggers
_fp_post_init_role_migration(env) on -u. post_init_hook only fires
on INSTALL in Odoo 19, not UPGRADE -- so Phase H's preview creation
wouldn't have auto-fired on entech without this script. Module
version bumped to 19.0.21.1.0 to match the migration directory.
C4: Team kanban template rewritten for Odoo 19 (<t t-name='card'> with
semantic <aside>/<main>) instead of legacy <t t-name='kanban-box'>.
Previous template threw 'Missing card template' at render.
IMPORTANT FIXES:
I1: SO state=sent Confirm button (id='action_confirm') now also gated
to group_fp_sales_manager. Previously only the state=draft button
was gated; Sales Reps could send-and-confirm via the secondary path.
I2: Designated Officials picker domain uses all_group_ids (transitive)
instead of group_ids (explicit only). Owner users now correctly
appear as eligible CGP DO candidates via the implied_ids chain.
I3: test_menu_visibility.py compliance hub xmlid corrected to
fusion_plating.menu_fp_compliance_hub (was
fusion_plating_compliance.menu_fp_compliance_hub which doesn't exist
-- the hub menu is defined in core's fp_menu.xml). Tests were
silently skipTest-ing.
I4: _inverse_plating_role chatter audit reads old role from DB via SQL
(bypasses cache) so 'old -> new' displays actual values, and
short-circuits no-op writes.
I5: _FP_ROLE_MAPPING_RULES reordered: cgp_designated_official fires
BEFORE admin/uid_1_or_2 so admin+DO users keep the capability_delta
marker that triggers res.company.x_fc_cgp_designated_official_id
auto-set during migration.
I6: _cron_purge_expired_migrations skips groups with active users
instead of unlink-ing unconditionally. Defense against rollback
safety being bypassed by manual role assignments post-migration.
CLAUDE.md updated with 3 new durable rules (13b kanban card template,
13c group_ids vs all_group_ids, 13d post_init_hook only on install).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Phase H of permissions overhaul (LAST subagent phase).
New models:
- fp.migration.preview (state: pending/approved/cancelled/rolled_back)
- fp.migration.preview.line (one per active internal user)
On -u, post_init_hook creates a preview in 'pending' state, walks all
active non-share users through the 12-rule mapping predicate chain
(first match wins, highest precedence first), and schedules a
mail.activity on every Owner.
Mapping table (per spec Section 5):
uid 1/2 / Administrator -> owner
CGP DO (existing) -> owner + res.company DO field set
CGP Officer -> quality_manager
Manager / Shop Mgr (old) -> manager
Accounting -> manager
Estimator-without-Manager -> sales_rep (flagged: loses confirm)
Supervisor / Receiving -> shop_manager
Operator -> technician
catchall -> 'no'
Owner clicks 'Approve & Run' on the preview form -> sudo write removes
old plating groups, adds new role's group, posts Markup chatter audit.
Optionally sets res.company.x_fc_cgp_designated_official_id for the DO.
30-day rollback window via JSON snapshot of groups_id per line. Daily
cron (Fusion Plating: Purge Expired Role Migrations) clears snapshots
+ unlinks old [DEPRECATED] groups after 30 days.
ACL: fp.migration.preview + .line both Owner-only (CRUD).
Menu: Plating > Configuration > Role Migrations (Owner-only).
Tests cover: only-Owner-can-approve, approve advances state, cancel
blocks after approval, rollback restores groups_id, Estimator warning
flagged, uid 2 maps to owner, rollback blocked after 30 days.
Per CLAUDE.md: ir.cron uses only Odoo-19-valid fields (no numbercall,
no doall). Post-init hook is idempotent — won't double-create previews
or re-fire if all users already migrated.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Phase G of permissions overhaul.
G2: sale.order.action_confirm now requires group_fp_sales_manager
(spec Section 2.B). Sales Reps can save drafts but cannot move SOs
to 'sale' state. UserError raised with clear message if attempted.
G3: Fixed audit-finding-11 typo bug in 2 files. The original code
checked has_group('fusion_plating.group_fusion_plating_administrator'),
an xmlid that has NEVER existed - so the gate always returned False
and only the Manager-side check actually fired. Fixed both:
- fusion_plating_invoicing/models/res_partner.py:34
- fusion_plating_configurator/wizard/fp_direct_order_wizard.py:467
Both now check has_group('fusion_plating.group_fp_manager') which
transitively includes Owner via implied_ids.
G4: Swept all Python has_group() calls to reference new group xmlids.
Backward-compat keeps old refs working today (Phase A's implied_ids),
but the sweep ensures correctness after the 30-day rollback window
deletes old groups. Replacements:
group_fusion_plating_operator -> group_fp_technician
group_fusion_plating_supervisor -> group_fp_shop_manager_v2
group_fusion_plating_manager -> group_fp_manager
group_fusion_plating_admin -> group_fp_owner
group_fusion_plating_cgp_officer -> group_fp_quality_manager
group_fusion_plating_cgp_designated_official -> group_fp_owner
group_fp_estimator -> group_fp_sales_rep
group_fp_accounting -> group_fp_manager
group_fp_receiving -> group_fp_shop_manager_v2
group_fp_shop_manager (legacy) -> group_fp_manager
G1: test_sales_manager_gate.py covers the new confirm gate (SR
blocked, SMg allowed, Manager allowed via diamond implication).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Phase F of permissions overhaul.
Adds res.users.x_fc_plating_role Selection field (8 options matching
the role hierarchy). Compute reads highest plating group from
groups_id (precedence: owner > QM > manager > sales_manager >
shop_manager > sales_rep > technician). Inverse uses sudo().write()
to clear all plating-role groups (additive-by-default m2m (3, id))
then adds the chosen one, and posts a Markup-wrapped chatter audit
naming the actor.
New Owner-only menu: Plating > Configuration > Team. Standard
res.users kanban grouped by x_fc_plating_role with records_draggable
for drag-and-drop role changes. Domain hides shared/portal users
and archived users.
res.company gains two Designated Official fields:
- x_fc_cgp_designated_official_id (CGP DO per Defence Production Act §22)
- x_fc_nadcap_authority_user_id (Nadcap signer)
Both tracking=True for audit. View-level domain restricts picker to
Owner or Quality Manager users via [(ref('...'), ref('...'))] xmlid
domains. New 'Plating Designated Officials' page on res.company form,
Owner-only visibility.
Tests in test_team_page.py cover compute/inverse/chatter/menu.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Phase E of permissions overhaul. The landing resolver now dispatches
based on the user's highest role (per spec Section 3):
Owner -> Manager Desk
Quality Mgr -> Quality Dashboard
Manager -> Manager Desk
Sales Manager -> Sale Orders
Shop Manager -> Plant Kanban (v2 layout) or Workstation (legacy)
Sales Rep -> Quotations
Technician -> Plant Kanban / Workstation
User override (x_fc_plating_landing_action_id) still wins; company
default and hardcoded Sale Orders are fallbacks. Layout-flag-aware via
ir.config_parameter['fusion_plating_shopfloor.layout'] (v2 vs legacy).
x_fc_pickable_landing field added to BOTH ir.actions.act_window AND
ir.actions.client (Manager Desk / Plant Kanban / Quality Dashboard
are client actions). Resolver helper polymorphically calls
_render_resolved() on either model.
Tagged 3 of 4 new actions pickable: Manager Desk, Plant Kanban,
Quality Dashboard. (action_fp_shopfloor_landing doesn't exist as an
XML record — it's a JS component name only; legacy layout falls
through to company default gracefully via raise_if_not_found=False.)
Tightened picklist domain to filter by user accessibility (Technician
no longer sees Manager Desk in the dropdown). New compute field
res.users.accessible_landing_action_ids runs check_access_rights on
each pickable action.
Tests in fusion_plating/tests/test_landing_resolver.py.
CLAUDE.md updated with two durable rules:
- x_fc_pickable_landing lives on BOTH act_window and actions.client
- Role-based dispatch precedence and helper API
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Phase D Task D5 of permissions overhaul. Adds explicit groups= to
form-level elements so non-matching roles don't even SEE the buttons
they can't use:
- SO Confirm button → group_fp_sales_manager (Sales Rep sees the SO
in draft but no Confirm button — matches model-level gate from Phase G)
- SO pricing fields (price_unit/subtotal/total/untaxed/tax) →
group_fp_sales_rep (Technician/Shop Manager don't see pricing if
they navigate to an SO)
- Partner Account Hold tab → group_fp_manager (was the fold-in
group_fp_accounting; the audit-finding-11 _administrator typo lives
in res_partner.py and is Phase G's fix)
- CAPA Close + all state-transition buttons → group_fp_quality_manager;
edit fields use readonly="not user_has_groups(...)" so Manager
retains read+comment per spec section 2.C
- Audit Start/Findings/Close buttons → group_fp_quality_manager
- AVL Approve/Suspend/Reinstate/Remove → group_fp_quality_manager
(model uses Suspend+Remove instead of spec's literal 'Disqualify';
both surfaces gated, semantics match)
- Customer Spec edit fields → readonly for non-QM (Manager keeps
read access per spec; only inputs lock)
- FAIR Approve/Reject buttons → group_fp_quality_manager (Submit-
for-Review and Reset stay open to whoever created the FAIR)
- Certificate Issue button — Strategy B chosen: single button hidden
when cert_type=nadcap_cert AND user is not QM. Cleaner than splitting
into two buttons; no separate action_sign exists on fp.certificate
(Issue is the sign+publish action). FAIR lives in its own model;
fp.certificate only has nadcap_cert as a special type. The ir.rule
from Phase C enforces model-level writes independently.
- CGP form buttons (7 view files: ai, controlled_good, psa,
receipt_shipment, registration, security_incident, visitor) →
group_fp_quality_manager on every action button
Defense in depth: ir.rules and ACLs (from Phases B + C) already
restrict model access. These view gates are the UI layer that
matches.
Concerns:
- Spec line 192 names 'sale.order view — x_fc_account_hold_override'
but no such field exists in the codebase. Closest practical match
was the partner-side Account Hold management tab, which already had
a group= attribute. Re-gated there; no SO-side field to gate.
- AVL model has no action_disqualify per spec; uses suspend+remove.
Both gated to QM.
- fp.certificate has no action_sign (only action_issue). FAIR's
approve/reject covers the FAIR side; nadcap-cert Issue covers the
cert side via Strategy B.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Implementer concern from D1-D4 dispatch: plan template referenced
menu_fp_sales_root / menu_fp_shopfloor_root but actual xmlids drop
the _root suffix. Tests were silently skipping.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Phase D Tasks D1-D4 of permissions overhaul. Adds explicit groups=
attributes to:
- 9 top-level Plating menus (matrix per spec Section 2.E)
- Quality submenus: Audits, Customer Specs, AVL → QM-only
- Compliance hub child submenus (CGP, General, Safety, Aerospace,
Nuclear) → QM-only
- Operations submenus: Maintenance, Move Log, Labor History → Shop
Manager+; Replenishment Suggestions → Manager+
Replaces fragile inheritance + action-ACL-based visibility with
explicit per-menu gates. Now every role's menu tree is deterministic.
Also adds fusion_plating/tests/test_menu_visibility.py — per-role
matrix tests using ir.ui.menu.search_count with the test user.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Implementer-flagged concern: plan template referenced fp.approved.vendor.list
but actual model id is fusion.plating.avl. Tests were silently skipping
instead of exercising the AVL split.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Phase C of permissions overhaul (spec Section 2.C).
Manager keeps reactive Quality (NCR/Hold/Check/Cert/RMA — already gated
via Phase B sweep). QM gains exclusive write/create/unlink on strategic
Quality records:
- fusion.plating.capa: Manager → read-only (1,0,0,0); QM → full
- fusion.plating.audit: same split (if model present)
- fp.approved.vendor.list: same split (if model present)
- fusion.plating.customer.spec: same split
- Doc Control models: same split
Plus FAIR/Nadcap cert restriction via two new ir.rule records on
fp.certificate:
- Manager: write/create/unlink on certs where cert_type NOT in
('fair', 'nadcap')
- QM: write/create/unlink on all certs (overrides via OR within group)
- Read access unchanged for both (perm_read=False on the rules)
Tests in fusion_plating/tests/test_quality_split.py verify each side
of the split. Models that may not exist on all DBs (audit, AVL) use
skipTest gracefully.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The Phase B plan (commit 8eb2c2de) listed 12 modules to sweep, but the
codebase has 13 more modules with ACL CSVs referencing the old role
group xmlids. Backward-compat (Phase A's implied_ids chains) keeps
these working today, but the old groups will be deleted after the
30-day rollback window — so the sweep must cover ALL modules with
plating-group ACL refs to avoid post-rollback breakage.
Sweeps: batch, bridge_documents, bridge_maintenance, bridge_mrp
(uninstalled but file present), bridge_quality (planned removal),
bridge_sign, compliance, culture (retired), kpi, logistics,
notifications, portal, reports.
Pattern matches the original sweep:
group_fusion_plating_operator → group_fp_technician
group_fusion_plating_supervisor → group_fp_shop_manager_v2
group_fusion_plating_manager → group_fp_manager
group_fusion_plating_admin → group_fp_owner
group_fp_accounting → group_fp_manager
group_fp_receiving → group_fp_shop_manager_v2
group_fp_estimator → group_fp_sales_rep
group_fp_shop_manager (legacy) → group_fp_manager
cgp_officer → group_fp_quality_manager
cgp_designated_official → group_fp_owner
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Phase B of permissions overhaul. Mechanical text replacement across
11 ir.model.access.csv files:
- group_fusion_plating_operator -> fusion_plating.group_fp_technician
- group_fusion_plating_supervisor -> fusion_plating.group_fp_shop_manager_v2
- group_fusion_plating_manager -> fusion_plating.group_fp_manager
- group_fusion_plating_admin -> fusion_plating.group_fp_owner
- group_fp_estimator (configurator)-> fusion_plating.group_fp_sales_rep
- group_fp_accounting -> fusion_plating.group_fp_manager
- group_fp_receiving -> fusion_plating.group_fp_shop_manager_v2
- group_fp_shop_manager (legacy) -> fusion_plating.group_fp_manager
- group_fusion_plating_cgp_officer -> fusion_plating.group_fp_quality_manager
- group_fusion_plating_cgp_designated_official -> fusion_plating.group_fp_owner
Backward-compat: old group xmlids still resolve (Phase A's implied_ids
chains keep old ACLs working for users still holding old groups).
This sweep ensures future-state correctness: when old groups are deleted
after the 30-day rollback window, ACLs continue resolving via the new
group xmlids.
Also adds fusion_plating/tests/test_acl_migration.py with sample-based
per-role access checks. The 2 CAPA tests are expected to fail until
Phase C implements the Manager/QM quality split.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Group-structure tests for Phase 1 permissions overhaul. Covers:
- All 7 new res.groups records present (8th role "No" is implicit)
- Owner transitively implies base.group_system + every old group
- Manager forms the diamond (implies both Shop Manager and Sales Manager)
- Sales and Shop branches remain orthogonal at the leaf (Tech != Sales Rep)
- uid 1/2 auto-assigned to Owner
- Sequence numbers unique (renders dropdown predictably)
- New groups imply old for backward-compat (30-day rollback safety)
- Cross-module backward-compat chain works (e.g., Owner -> CGP DO)
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The previous commit (a53b0326) added implied_ids in fp_security_v2.xml
that referenced 5 xmlids from downstream modules (configurator/receiving/
invoicing/cgp). Since fusion_plating is the BASE module and loads first
at fresh install, those refs raised External-ID-not-found at install.
Fix: relocate the 5 cross-module implications into each downstream module's
own security file via additive (4, ref()) writes to the core group's
implied_ids. Odoo's XML data loader treats these as additive updates so
they stack cleanly across install + -u cycles.
Also: drop redundant <data noupdate="0"> wrapper in fp_security_v2.xml
to match sibling fp_security.xml's bare <odoo> shape.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Phase A of permissions overhaul (see docs/superpowers/specs/2026-05-23-*).
New groups (technician/sales_rep/shop_manager_v2/sales_manager/manager/
quality_manager/owner) defined in fp_security_v2.xml with implied_ids
chains that include old groups for backward-compat during 30-day rollback
window. Old groups display as [DEPRECATED] in user form.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Spec describes consolidation of 12 res.groups into 8 roles (No / Technician /
Sales Rep / Shop Manager / Sales Manager / Manager / Quality Manager / Owner),
role-based landing-page defaults, Owner-only Team management page, and
dry-run + Owner-approval migration workflow.
Plan breaks the work into 9 phases (A through I), ~40 TDD tasks, with
explicit file lists and entech deploy commands.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Two fixes from live testing of the 2026-05-24 redesign:
1. Job Workspace Back button routed to deprecated component.
onBack() hardcoded tag: 'fp_shopfloor_landing' so tapping a card on
the new plant kanban -> opening the workspace -> clicking Back
dropped the user into the OLD per-step kanban (the legacy OWL
component the data-record redirects don't reach because doAction
bypasses the data layer).
Fix: change the hardcoded tag to 'fp_plant_kanban'. Grep
confirmed it's the only such reference in JS.
2. Logo frame shape — wider, shorter, logo bigger.
140x140 square -> 280x110 rectangle. Better fit for horizontal
company logos (mark + name + tagline laid out left-to-right).
Uniform 18px padding on all sides so the image breathes evenly.
Image area is ~244x74 (vs old ~104x104), so a typical horizontal
logo renders ~50% wider. border-radius 28->22 for the flatter
rect; letter-mark placeholder font 52->44 to fit the shorter
frame.
Also augmented CLAUDE.md 'Legacy-action redirect' rule with a new
'grep JS for hardcoded doAction' clause — the XML-record redirect
trick only covers ir.actions.client data; OWL components with inline
this.action.doAction({tag: ...}) calls bypass the data layer entirely
and need a separate sweep.
Asset cache cleared (3 stale attachments).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
User feedback after live testing: the 84px logo frame felt too small
and the image inside used only a fraction of the frame area.
Bumped the frame to 140px (1.67x) — image scales with the container
via the existing max-width/max-height: 100% rule. Proportional
adjustments to padding (14→18), border-radius (20→28), margin-bottom
(12→16), and the letter-mark placeholder font (32→52).
SCSS-only change. Asset cache cleared (3 stale attachments).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
LS-T2..T6 of the tablet lock-screen redesign (LS-T1 backend shipped
separately in c6137100).
Files:
- _tablet_lock_tokens.scss (new — design tokens, dark/light branches
via $o-webclient-color-scheme, registered
first in manifest per project rule 8)
- tablet_lock.scss (full rewrite — gradient bg, glassmorphic
tiles, 4 entrance keyframes, hover lift,
click press, clocked-in pulse,
prefers-reduced-motion gate)
- tablet_lock.xml (extended — logo + clock + prompt blocks
wrapping the existing tile loop; tile
inner shape updated for avatar gradient,
has_photo conditional, is_clocked_in
modifier)
- tablet_lock.js (extended — state.clockText / dateText /
company, setInterval(60s) clock tick,
_formatTime / _formatDate / tileStyle /
avatarClass helpers per project rule 20)
- __manifest__.py (19.0.31.0.0 -> 19.0.32.0.0, registered
new tokens SCSS BEFORE tablet_lock.scss)
Verified live on entech:
- Module upgrade clean, registry loaded in 15.5s
- 6 stale asset attachments cleared
- Helpers in tablet_controller.py emit company payload + initials +
gradients correctly (Garry Singh -> GS, EN Tech -> ET, uid=5 ->
pink gradient)
- res.company.logo present (has_logo: True)
- All animations gated by prefers-reduced-motion per spec §6
CLAUDE.md updated with new Critical Rule 22 about Odoo 19 HTML fields
auto-wrapping plain-string writes — caught during Task 1 testing when
the original 'tagline equality' test failed because res.company.report
_header is an HTML field that wraps writes with <p> tags.
Closes the 6-task plan in
docs/superpowers/plans/2026-05-24-tablet-lock-screen-redesign-plan.md
Spec: docs/superpowers/specs/2026-05-24-tablet-lock-screen-redesign-design.md
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
LS-T1 of the tablet lock-screen redesign.
Adds 3 module-level helpers in tablet_controller.py:
_initials_from(name) — first/last initials for letter-mark fallback
_avatar_gradient_for(uid) — deterministic per-user color (8 gradients)
_lock_company_payload(env) — company name + tagline + logo URL block
Endpoint /fp/tablet/tiles now returns:
{ok, company:{id,name,tagline,logo_url,has_logo,initials},
tiles:[{user_id, name, initials, avatar_url, has_photo,
avatar_gradient, is_clocked_in, has_pin}, ...]}
Tagline reuses res.company.report_header (the existing invoice-letterhead
field) — no new model field. Falls back to 'Shop Floor Terminal' when
empty.
10 tests pass (initials edge cases, gradient determinism, payload shape).
The 'tagline matches input string' assertion was intentionally NOT added
— see new CLAUDE.md Critical Rule 22 about Odoo 19 HTML field
auto-wrapping that makes such an equality test brittle.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Hybrid Industrial Bold + Premium Glassmorphism direction approved
during brainstorming. Adds company branding (logo from
res.company.logo with letter-mark fallback), real-time clock, tighter
3-column tile grid for ~10-15 operator small shops, dual dark/light
mode via compile-time $o-webclient-color-scheme branch, 7-animation
catalogue gated by prefers-reduced-motion.
Backend touch: extend /fp/tablet/tiles payload with company block +
per-tile initials/avatar_gradient/has_photo. Two small helper
functions in tablet_controller. No DB migration.
Frontend touch: new _tablet_lock_tokens.scss (loads first), full
rewrite of tablet_lock.scss, extend XML + JS for clock + company.
Mockup: .superpowers/brainstorm/1983-1779585812/content/lock-final.html
(in-repo since the brainstorm session used --project-dir).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
User feedback after live testing: cards were too cramped on the 9-column
board. Restoring the Variant C mockup proportions and letting the board
scroll horizontally on smaller viewports (user explicitly accepted
side-scrolling).
Changes:
- .board grid: repeat(9, 1fr) → repeat(9, minmax(300px, 1fr))
plus overflow-x: auto. Each column ~300px so the card has room to
breathe. ~6 columns visible on 1920px desktop, ~4 on 1366px tablet,
smooth horizontal scroll for the rest.
- .col-scroll: gap 4→8, max-height eased so cards aren't packed.
- .o_fp_plant_card: padding 8/10→12, gap 4→6, base font 11→12.
- card-wo: 13→16 (matches mockup header size).
- card-step: 12→14.
- chips: padding 1/6→2/8, font 10→11, radius 10→12.
- mini-timeline blocks: 8→16px tall (current step 11→22px), labels
8→9px. Matches the mockup's punchy timeline strip.
- progress bar: max-width 60→100, height 3→4.
- operator pill / icon-row: bumped to match card scale.
No backend changes. SCSS-only. Asset cache cleared (3 attachments).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The plant-view rollout left two legacy ir.actions.client data records
still claiming tag='fp_shopfloor_landing':
- action_fp_plant_overview (Plant Overview)
- action_fp_shopfloor_tablet (Shop Floor — Tablet Station)
The landing-action resolver dispatched the new view correctly when the
user clicked the Plating root menu, but bookmarks / breadcrumbs /
QR-scan landings / direct URLs still routed through these legacy
actions and loaded the per-step kanban (OWL component is still
registered for back-compat).
Flipping their tag to fp_plant_kanban means every entry point now
opens the new view. The legacy fp_shopfloor_landing OWL component
stays registered (no code removed) but no XMLID points at it
anymore — safe to delete in a future cleanup.
Also documented this as a durable convention in CLAUDE.md under
'Legacy-action redirect (general rule for OWL component swaps)'.
Verified on entech:
- action 1129 (Shop Floor) tag: fp_shopfloor_landing → fp_plant_kanban
- action 1133 (Plant Overview) tag: fp_shopfloor_landing → fp_plant_kanban
- 3 stale asset bundles cleared
- Module re-upgraded clean, registry rebuilt in 15.7s
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
PV-Phase5 of the plant-view redesign. Final phase — flips the default
of x_fc_shopfloor_layout from 'legacy' to 'v2' and updates CLAUDE.md
with the new architecture rule.
Verified on entech:
- HTTP 200 on /web/login
- Shopfloor module loads cleanly with all 19 new frontend files
- /fp/landing/plant_kanban returns the assembled payload with 9
columns + denormalized cards
- Card state distribution: 22 contract_review + 8 no_parts + 1 running
(sample data only — dev system)
- Asset bundle re-compiled (9 stale attachments cleared)
- ir.config_parameter['fusion_plating_shopfloor.layout'] = 'v2' set
To switch back to legacy: Settings → Fusion Plating → Shop Floor
Layout, or UPDATE ir_config_parameter SET value='legacy' WHERE
key='fusion_plating_shopfloor.layout'.
CLAUDE.md gets a new ~80-line section documenting:
- Why the redesign (per-step kanban produced duplicate cards)
- 9-column layout + step-kind → area mapping (spec D3, D4, D5)
- 13-state catalog + precedence dispatch in _compute_card_state
- Backend single-endpoint payload shape (/fp/landing/plant_kanban)
- Frontend OWL component tree + critical implementation gotchas
(rule 20 OWL scope, rule 8 SCSS @import, dark-mode compile-time)
- How to switch back to legacy
Closes the 20-task plan in
docs/superpowers/plans/2026-05-23-shopfloor-plant-view-plan.md
Spec: docs/superpowers/specs/2026-05-23-shopfloor-plant-view-design.md
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
PV-Phase4 of the plant-view redesign. 19 new files implementing the
6-component OWL tree plus design tokens.
Components (each = JS + XML + SCSS triple):
- FpMiniTimeline — 9-step bar consuming mini_timeline_json
- FpPlantCard — Variant C card; 13 state-* CSS classes; tap
opens fp_job_workspace
- FpColumnHeader — column label + count badge + 'You're here'
badge when paired
- FpKpiTile — clickable KPI button with urgent/warn/good
variants and active state
- FpFilterChip — toggleable chip
- FpPlantKanban — top-level orchestrator: 10s polling, mode
toggle, search + 6 filter chips, board with
9 fixed columns, localStorage filter persistence
SCSS:
- _plant_tokens.scss (loads first, exposes $plant-* vars to every
later file — required because Odoo 19 forbids @import in custom
SCSS, manifest order IS the concat order)
- Dark mode via $o-webclient-color-scheme compile-time branch
Manifest registers all assets in dependency order: tokens → component
SCSS → component XML → leaf JS → top-level JS. Mirrors the existing
project pattern.
Critical patterns honored:
- Project rule 20 (no String/Number/parseInt in OWL templates):
all coercion in JS, string literals in foreach arrays.
- No t-out without markup() (none in this batch — all card text is
pre-formatted by the controller).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
PV-Phase3 of the plant-view redesign.
- /fp/landing/plant_kanban JSONRPC endpoint returns {kpis, columns,
cards} in one payload. One card per fp.job; cards denormalized so
the OWL component doesn't fan out RPCs. Server-side filter handling
for All / Mine / Running / Blocked / Overdue / FAIR. Within-column
sort by (overdue, _SORT_PRIORITY[card_state], due_date).
- fusion_plating_shopfloor.action_fp_plant_kanban client action
registered alongside the existing fp_shopfloor_landing action.
- fp_landing_data.xml resolver extended to read the layout flag and
dispatch to v2 when x_fc_shopfloor_layout='v2' (default still legacy).
Card payload (23 fields): WO, customer, PN+rev, qty, PO, recipe, spec,
tags, current step + work centre, state chip, mini_timeline, operator,
icons (signoff / bake / tracking / etc.), progress.
State-chip mapping per spec §6.1 — one map keyed by card_state with
running-time elapsed, idle-hours, and operator-name interpolation.
Verified live — card payload sample on WO-30036 (contract_review state)
produces all expected keys + 9-element mini_timeline.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
PV-Phase2 of the plant-view redesign.
Implements the 13-state classifier on fp.job:
- card_state Char field, stored + indexed for fast filtering
- _compute_card_state with explicit precedence dispatch matching
spec §6.2 / §9.3 exactly (no_parts → on_hold → awaiting_signoff
→ awaiting_qc → bake_due → predecessor_locked → idle_warning →
done → contract_review → running/_mine → ready/_mine)
Six precedence helpers, each isolated for testability:
_fp_inbound_not_received, _fp_has_open_hold, _fp_has_pending_qc,
_fp_bake_window_due_soon, _fp_is_mine + _fp_has_unfinished_predecessors
on fp.job.step.
mini_timeline_json compute: 9-element array (one per column) with
state in {done, current, upcoming} and an optional 'variant' on the
current marker keyed to card_state for renderer color mapping.
Verified live:
- 14 jobs in contract_review (no active step yet)
- 8 in no_parts (confirmed + draft fp.receiving)
- 1 running (WO-30051 with Pre-Measurements at Plating column)
- mini_timeline JSON renders the full 9-area structure with the
plating slot marked current+variant=running.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Consolidated commit of session work already deployed to entech and
verified via the deep audit + the persona walk:
S22 — Signoff gate (fp.job.step.requires_signoff was 100% unenforced,
42/42 done steps had NULL signoff_user_id). Three-piece fix:
_fp_autosign_if_required (captures finisher on button_finish),
_fp_check_signoff_complete (raises UserError if NULL after autosign),
action_signoff (explicit supervisor pre-sign). Bypass:
fp_skip_signoff_gate=True.
S23 — Transition-form gate (same dormant-field shape as S22, caught
preventively before recipe authors flipped requires_transition_form
on). Model helpers on fp.job.step.move + controller gate in
move_controller (parts commit) + pre-reject in rack commit.
F7 — Chatter standardization: _fp_create_qc_check_if_needed,
_fp_fire_notification, _fp_create_delivery silent failures now also
post to job chatter instead of only logging to file.
UI fixes:
- Critical Rule 20 documented + applied: OWL templates only expose
Math as a global. Calling String(d) inside t-on-click throws
'v2 is not a function'. Fixed pin_pad.xml (string array instead of
number array with String() coercion). Also swept parseInt/
parseFloat in recipe_tree_editor + simple_recipe_editor.
- Notes panel HTML escape fix: chatter messages off /fp/workspace/load
were rendered via t-out, escaping the HTML. Wrap with markup() in
job_workspace.js refresh() before assigning to state.
Versions:
fusion_plating 19.0.20.8.0 → 19.0.20.9.0
fusion_plating_jobs 19.0.10.20.0 → 19.0.10.23.0
fusion_plating_shopfloor 19.0.30.2.0 → 19.0.30.5.0
All deployed to entech (LXC 111) and verified live.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Adds the Phase 6.3 fpRpc wrapper to the web.assets_backend bundle.
Placed before its consumers so the `import { fpRpc } from "./services/fp_rpc"`
calls in job_workspace, shopfloor_landing, manager_dashboard, and
hold_composer resolve.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
JobWorkspace, ShopfloorLanding, ManagerDashboard, and the embedded
FpHoldComposer now call fpRpc() for write-path endpoints (start/finish
step, hold create, sign-off, milestone advance, work-centre move,
assign-worker, assign-tank, manager takeover). fpRpc auto-injects
tablet_tech_id from the tech_store so the server can rebind env via
env_for_tablet_tech() and credit the right user.
Read-path RPCs (workspace/load, landing/kanban, manager/overview,
manager/funnel, manager/approval_inbox, manager/at_risk, shopfloor/scan)
stay as plain rpc() — no audit benefit, no need for the extra plumbing.
Also wires tablet_tech_id into /fp/shopfloor/plant_overview/move_card
which I missed in P6.3.3 — surfaced when grepping JS for write callers.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
10 endpoints in shopfloor_controller (log_chemistry, start_bake, end_bake,
start_wo, stop_wo, bump_qty_done, bump_qty_scrapped, log_thickness_reading,
quality_hold, mark_gate) and 3 in manager_controller (assign_worker,
assign_tank, take_over) now accept a `tablet_tech_id` kwarg. Each rebinds
env via env_for_tablet_tech() so writes carry the correct uid even when
the OS session belongs to the persistent tablet user.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
hold, sign_off, advance_milestone each accept tablet_tech_id and
rebind env via env_for_tablet_tech. Writes (Hold.create, button_finish,
action_advance_next_milestone) now carry the tech-of-record's uid.
load endpoint is read-only and untouched.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Client-side fpRpc() is a drop-in for rpc() that automatically injects
tablet_tech_id from the tech_store into every action call. Read-only
endpoints can keep using plain rpc().
Server-side env_for_tablet_tech(env, tablet_tech_id) returns an env
scoped via with_user() when the id is a valid active user; otherwise
returns the original env unchanged. Controllers call this at the top
of action methods so all subsequent writes carry the right uid.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Registers fp_tablet_pin_setup as an ir.actions.client tag. Triggered
from res.users preferences via action_open_tablet_pin_setup (added
to res_users.py in P6.1.1). Three-stage flow:
loading → check if user has existing PIN via search_count
old → enter current PIN (skipped if first-time)
new → choose new PIN
confirm → enter new PIN again
done → success toast + auto-close 1.5s later
Each stage reuses FpPinPad with a different onSubmit + title. On
mismatch / server error, resets to the first stage with a notification.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Three OWL client actions all wrap their root in <FpTabletLock>:
ShopfloorLanding wraps o_fp_landing
JobWorkspace wraps o_fp_ws
ManagerDashboard wraps o_fp_manager
Each adds FpTabletLock to static components, imports tech_store, and
gains a handOff() method that calls techStore.lock(). The Hand-Off
button (yellow, lock icon) lands next to the scan/QR controls in each
header — pressing it instantly returns the tablet to the tile grid
without waiting for the idle timer.
Component composition (per spec §6.5):
FpTabletLock
if isLocked → tile grid + FpPinPad
else → existing client action (via <t t-slot="default"/>) + FpIdleWarning
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Top-level wrapper that renders lock screen (tile grid + PIN pad) when
no tech is signed in, and renders <t t-slot="default"/> otherwise.
Drives the auto-lock countdown via the activity_tracker service +
sends a /fp/tablet/ping heartbeat every 60s while a tech is signed in.
Tiles fetch from /fp/tablet/tiles using the localStorage station id
(set by ShopfloorLanding on QR pair / station picker selection).
State machine for the lock screen body:
loadingTiles → tiles list → tile tapped → PinPad → unlock RPC
↑
onPinCancel → back to tiles
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Fixed-position yellow-border overlay + countdown toast shown during
the last N (default 30) seconds before auto-lock. Pure props-driven —
secondsRemaining is the only input; parent (FpTabletLock) decides
when to mount and unmount. Box-shadow pulse animation runs CSS-only
so OWL doesn't need to re-render every tick.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Reusable 4-digit PIN pad. Auto-submits on the 4th digit via the
onSubmit prop. On wrong PIN, shake animation + dots clear + error
banner (caller controls the message via the returned {ok:false, error}).
Used by FpTabletLock (unlock flow) and FpPinSetup (set/change flow).
Dark-mode SCSS branch follows the same $o-webclient-color-scheme
pattern as the rest of the shopfloor components.
Also registers tech_store + activity_tracker services in the asset
bundle (assets/web.assets_backend) before the pin_pad files, since
the pin_pad/tablet_lock components consume them.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Two registry-level services:
tech_store Shared reactive state holding currentTechId after a
successful PIN unlock. Other components subscribe via
useService("fp_shopfloor_tech_store") and read
currentTechId to inject into action RPCs. setTech(id, name)
on unlock; lock() on auto-lock / Hand-Off.
activity_tracker Document-level event tracker for pointerdown / touchstart
/ keydown / visibilitychange. Mouse-move alone deliberately
EXCLUDED — a tool resting on a tablet would otherwise keep
the session alive indefinitely. Public API:
bump(), getSecondsUntilLock(), getWarnThresholdSec()
Reads thresholds from ir.config_parameter at start +
every 5 min (so manager edits propagate within a shift).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Two view inheritances on res.users:
(a) Preferences form — adds a 'Tablet PIN' group with a 'Set / Change
Tablet PIN' button that triggers action_open_tablet_pin_setup → the
fp_tablet_pin_setup OWL client action (Phase 6.2). Shows PIN Last
Set as read-only context.
(b) Standard res.users form — header button 'Reset Tablet PIN' visible
only to the fusion_plating manager group; hidden when no PIN is set
(via the set_date invisible field reference). Confirms before clearing.
Calls the clear_tablet_pin method from the model.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Adds two fields to fusion.plating.shopfloor.station:
- x_fc_authorised_user_ids (Many2many → res.users): restricts the
tablet lock-screen tile grid to a specific roster per station.
Empty = all operator-group users shown.
- x_fc_idle_lock_minutes (Integer, nullable): per-station override
for the auto-lock idle threshold; null = use system parameter.
Plus data/fp_tablet_config_data.xml registers four ir.config_parameter
defaults (noupdate=1 — manager can override via Settings → Technical
→ Parameters):
fp.shopfloor.tablet_idle_lock_minutes = 5
fp.shopfloor.tablet_pin_fail_threshold = 5
fp.shopfloor.tablet_pin_fail_lockout_minutes = 5
fp.shopfloor.tablet_warn_seconds_before_lock = 30
Form view surfaces both new fields in a dedicated 'Tablet PIN Gate'
group.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Tiles returns the lock-screen grid: operator-group users, sorted
clocked-in-first then alphabetical, with avatar URL + has_pin flag.
Honours station.x_fc_authorised_user_ids when non-empty (Phase 6.1.6
adds that field). Ping is a lightweight ack used by FpTabletLock as
a heartbeat — logs current_tech_id at DEBUG for forensic visibility.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Verifies PIN, resets failure counter on success, increments + locks out
on 5 consecutive failures (configurable via ir.config_parameter
fp.shopfloor.tablet_pin_fail_threshold + tablet_pin_fail_lockout_minutes,
both defaulting to 5).
Returns informative payloads:
ok=true current_tech_id, current_tech_name
needs_setup=true user has no PIN yet
locked_until lockout in effect (rejects even correct PIN)
attempts_remaining failed but not yet locked
Logs INFO on success, WARNING on failure (with running counter +
locked flag).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
set_pin is self-service: requires old PIN if a hash exists, validates
4-digit format. reset_pin_for is manager-only (enforced server-side
via has_group); clears the hash + posts to chatter.
Both endpoints log INFO on success and WARNING on access-control denials.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
PBKDF2-SHA256 + 16-byte salt + 200k iterations on res.users. Format
of the stored hash string is <salt_hex>$<digest_hex>. Field is
manager-readable only (groups=group_fusion_plating_manager); helpers
that need to read or write it use .sudo() internally so operator-level
callers can still set/verify their own PIN.
Adds set_tablet_pin / verify_tablet_pin / clear_tablet_pin model
methods + action_open_tablet_pin_setup that triggers the OWL setup
modal (Phase 6.2). Tests cover hash uniqueness, verify, clear with
chatter post, and the 4-digit format guard.
Tests verified on entech: -u fusion_plating_shopfloor --test-tags fp_tablet_pin
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Sequel to the 2026-05-22 tablet redesign (Phases 1-5). Adds a tile-grid
lock screen + 4-digit PIN per tech + 5-min auto-lock + audit propagation
so multiple techs sharing one tablet get correctly-attributed actions.
Key design choices:
- 4-digit PIN (industry norm), PBKDF2-SHA256 with 200k iterations
- Per-user lockout after 5 failures (not per-tablet)
- Single Odoo session + tablet_tech_id kwarg for audit (no JS reload on
every tech switch)
- Manager-side reset only (no SMS/email infra)
- Server-side step timer keeps running on lock (auto-pause cron is
the upper-bound safety net)
Three sub-phases (6.1 backend / 6.2 frontend lock / 6.3 audit kwarg
propagation), each independently deployable.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Phase 3 removed the menu_fp_shopfloor_plant_overview menuitem from
fp_menu.xml, but Odoo doesn't auto-delete orphan records when XML
disappears — the menu stayed in the database. Combined with P3.5's
action retarget (action_fp_plant_overview tag → fp_shopfloor_landing),
clicking it landed on the same Landing component as Workstation —
hence the duplicate menu items both opening the same screen.
Adds <delete model='ir.ui.menu' id='...'> in legacy_menu_hide.xml so
future -u runs scrub the orphan. Drops the now-defunct group_ids
block for the deleted menu. The action record stays (bookmark
back-compat).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The Phase 4 endpoints (/fp/manager/funnel, approval_inbox, at_risk)
all use fields.Datetime.now() but the controller only imported http
+ request. Hitting the Workflow Funnel tab on Manager Desk threw:
NameError: name 'fields' is not defined
Funnel auto-loads on dashboard mount → infinite spinner + 'Funnel:
Odoo Server Error' notification. Same bug would have hit at_risk
and approval_inbox on first navigation.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Same regression as the previous commit — b0070afc removed all 4 quick_look
related fields, my first fix only caught 2 of them. Restoring the remaining
2 so the quick-look view fully validates.
Commit b0070afc removed these two related fields from fp.job.step but
the view fp_job_step_quick_look_views.xml still references them. The
mismatch was dormant because entech never ran -u between b0070afc
and the 2026-05-22 deploy. Re-running -u during the Phase 1-4 deploy
caught it:
Field "quick_look_part_catalog_id" does not exist in model
"fp.job.step"
Restoring both as related fields (zero-cost, fixes the view without
touching XML).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Caught during entech deploy of the Phase 2 auto-pause cron. Odoo 19
ir.cron no longer accepts numbercall or doall fields; the load fails
with:
ValueError: Invalid field 'numbercall' in 'ir.cron'
Removed both from ir_cron_autopause_stale_steps. The other crons in
the same file (nudge stale paused / in_progress) already used the
minimal field set — matching that pattern now.
Also added a CLAUDE.md section so future-Claude doesn't reintroduce
the speculative fields.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Replaces the stale "Plant Overview Dashboard" section with a current
"Shop Floor Architecture" section covering the Phase 1-4 deliverables:
- 3 OWL client actions (Landing / JobWorkspace / Manager Dashboard)
- 5 shared OWL services
- Backend endpoints (workspace / landing / manager)
- Auto-pause cron config knob (ir.config_parameter name)
- Key new model fields with their purpose
- Operator ACL lift summary
- Deprecated-but-still-live legacy surfaces (Phase 5 cleanup pending)
- Old patterns to avoid
Links to the spec + plan docs as the authoritative reference.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Plan tasks P4.5 through P4.10 batched. Existing 3-column Plant Board
becomes one tab of four; adds Workflow Funnel (default), Approval
Inbox, and At-Risk siblings. Adds 2 new KPI tiles for Pending Cert +
At-Risk.
WORKFLOW FUNNEL (default tab)
Calls /fp/manager/funnel. Renders one row per fp.job.workflow.state
with stage chip + count + top 5 WO cards. Tap a card → JobWorkspace.
Bar chart bar behind each row scales with stage count.
APPROVAL INBOX
Calls /fp/manager/approval_inbox. Three strips: Holds to Release,
Certs to Issue, Scrap to Review. Per-row open + Open Workspace
buttons. Tab badge shows total pending count.
PLANT BOARD (existing — relocated as one tab)
The 3-column Needs Worker / In Progress / Team layout that already
exists, wrapped in t-if="activeTab === 'plant_board'". No behaviour
change — still uses /fp/manager/overview with 8s refresh.
AT-RISK
Calls /fp/manager/at_risk. 3 sub-panels: Trending Late (sorted by
late_risk_ratio desc), Hold Reasons (read_group), Bottleneck heatmap
(bottleneck_score from P4.1 with red/yellow/green bars).
KPI STRIP (new conditional tiles)
Pending Cert — count from inbox.certs_to_issue, click to open Inbox tab.
At-Risk — count from at_risk.trending_late, click to open At-Risk.
Auto-refresh: 8s for /fp/manager/overview (existing); the active tab's
data also refreshes every 8s via refreshActiveTab().
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Plan tasks P4.2 + P4.3 + P4.4 batched. Adds the backend data layer
for the Manager Desk's 3 new sibling tabs (Phase 4 tablet redesign).
POST /fp/manager/funnel
Workflow funnel: jobs grouped by fp.job.workflow.state. Returns
stages[] with count + top 5 WO cards per stage. Drives the
default tab on the refactored dashboard.
POST /fp/manager/approval_inbox
Four buckets: holds_to_release (state=on_hold|under_review),
certs_to_issue (all_steps_terminal + draft cert), scrap_to_review
(last 24h mark_for_scrap holds), override_requests (deferred —
empty placeholder).
POST /fp/manager/at_risk
Three panels: trending_late (top 20 by late_risk_ratio desc),
hold_reasons (read_group on hold_reason), bottleneck (top 10
work centres by bottleneck_score from P4.1).
All endpoints respect optional facility_id scope. Cheap implementations
— no caching yet; performance can be added if entech load demands.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Computes for the Manager At-Risk heatmap (Phase 4 tablet redesign).
Non-stored — recomputed on /fp/manager/at_risk read; that endpoint
caches its full payload for 60s so the cost is bounded.
bottleneck_score = active_step_count * avg_wait_minutes
avg_wait_minutes = rolling-7-day avg of (date_started - create_date)
Work centres with high score show red in the heatmap — combination
of queue length AND average wait time.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Plan task P3.6 — pragmatic deviation. The plan called for stubs that
internally route to /fp/landing/kanban + reshape; in practice the
legacy fp_shopfloor_tablet OWL component (still registered, just
unhooked from the menu) consumes a much richer payload (my_queue,
active_wo, baths, bake_windows, gates, holds, pending_qcs, stations)
than /fp/landing/kanban returns. Gutting tablet_overview to a stub
would break that legacy component.
Instead: add explicit DEPRECATED markers + INFO log lines on the three
endpoints (tablet_overview, plant_overview, queue). Bodies stay intact
so the legacy components keep working until Phase 5 cleanup retires
both endpoints AND the legacy OWL components together.
Note: /fp/shopfloor/plant_overview/move_card is NOT deprecated — the
new Landing component still uses it for drag-and-drop.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Plan task P3.5. Single 'Workstation' menu item replaces both the
legacy 'Tablet Station' and 'Plant Overview' entries. The new
fp_shopfloor_landing component has a Station/All-Plant toggle so
one menu covers both old surfaces.
Old action records redirected for back-compat (so existing bookmarks
+ smart-button references keep working):
action_fp_shopfloor_tablet tag → fp_shopfloor_landing
action_fp_plant_overview tag → fp_shopfloor_landing
params → {'mode': 'all_plant'}
The legacy OWL components (fp_shopfloor_tablet, fp_plant_overview)
remain registered — no code removed, just no menu points at them.
Phase 5 cleanup will remove the OWL components after a release of
soak time on entech.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Plan tasks P3.2 + P3.3 + P3.4 batched. Full ShopfloorLanding OWL
client action — replaces fp_shopfloor_tablet AND folds in
fp_plant_overview.
Header strip Title, station chip, station picker dropdown,
Station/All-Plant mode toggle, QR scan controls,
last-refresh indicator.
KPI strip 4 tech-relevant tiles: Ready · Running ·
Bakes Due (warning) · Holds (red when > 0).
Search Live debounced (200ms) across WO# + customer +
part. ESC clears.
Kanban board Columns = work centres from /fp/landing/kanban.
Cards = FpKanbanCard (Phase 1 — P1.7).
Drag-and-drop reuses existing
/fp/shopfloor/plant_overview/move_card.
Card tap doAction → fp_job_workspace with
{job_id, focus_step_id}.
QR scan FP-STATION pairs, FP-JOB / FP-STEP jump to the
Workspace.
Mode + station_id persist in localStorage (LS_STATION_ID, LS_MODE).
Auto-refresh every 15s; suppressed during a drop and for 5s after.
Registers client action `fp_shopfloor_landing`. Menu rewire + endpoint
stubs land in P3.5 + P3.6.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Plan task P3.1. New JSON-RPC endpoint for the Shop Floor Landing
client action (Phase 3). Two modes:
station — paired WC + Unassigned + next 1-2 WCs in recipe flow
all_plant — every active WC, recipe-flow order (replaces the data
path for the standalone fp_plant_overview action)
Returns {columns: [{work_center_id, work_center_name, cards}], kpis:
{ready, running, bakes_due, holds}, stations: [...], facility_name,
server_time}. Card payload matches the KanbanCard OWL component
(P1.7) — same shape, no client-side adapter needed.
Light implementation — no urgency scoring or batch prefetch yet.
Both can be ported from plant_overview if performance demands.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Plan task P2.6. Per the spec's "techs wear multiple hats" rule, lift
gates so technicians can do their work without permission walls:
fp.certificate operator: read → read+write
(flip draft→issued from tablet)
fp.thickness.reading operator: read → read+write+create
(capture Fischerscope readings from tablet)
fp.job.node.override operator: NEW read-only
(see opt-out badges on steps)
Supervisor-only operations (step Skip, hold Release, override
Re-include) remain enforced in workspace_controller, not ACL — so the
ACL stays minimal and the controller centralizes the gate logic.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Plan tasks P2.4 + P2.5 batched.
Adds _cron_autopause_stale_steps method on fp.job.step + 30-min cron
registration. Flips in_progress steps idle > threshold to paused with
a chatter audit ("Auto-paused after Nh idle. Resume from the tablet
when work continues.").
Threshold from ir.config_parameter:
fp.shopfloor.autopause_threshold_hours (default 8.0)
Recipe nodes opt out via fusion.plating.process.node.long_running
(added in P2.1) — useful for 24h bakes and multi-shift soaks.
Fixes the 411-hour ghost timer that motivated the redesign. Doesn't
replace the existing nudge crons — those still notify the supervisor;
this one actually pauses the timer.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Plan tasks P2.2 + P2.3 batched (both small additive computes on fp.job;
local tests not run between them — entech verifies).
late_risk_ratio — stored Float, remaining_planned / minutes_to_deadline.
Drives the Manager At-Risk view (Phase 4).
Recomputes on step state, duration, deadline changes.
active_step_id — non-stored Many2one. Currently in_progress step
(lowest sequence if multiple — defensive).
Drives JobWorkspace landing focus.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Plan task P2.1. Boolean on fusion.plating.process.node that exempts
steps generated from this node from the shop-floor auto-pause cron
(added in P2.4/P2.5). Use for 24h bakes, multi-shift soaks, and
similar long-but-legitimate operations.
Toggle visible on the process-node form for operation/step types,
grouped with parallel_start in the Behaviour section.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Plan task P1.16. Header button on the fp.job form that opens the
JobWorkspace OWL client action focused on the current WO. Primary
entry point for techs before the Landing kanban (Phase 3) ships;
remains as a back-office shortcut after.
Hidden when state == 'draft' (no steps to work yet).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Plan tasks P1.8 through P1.11 batched into one commit (local tests not
run between them; entech is the verification env).
POST /fp/workspace/load — full payload for one fp.job
POST /fp/workspace/hold — quality.hold create with photo
POST /fp/workspace/sign_off — signature + finish step atomic
POST /fp/workspace/advance_milestone — fire next_milestone_action
Each endpoint logs INFO on success, EXCEPTION on failure, returns a
consistent {'ok': bool, 'error': str?} envelope. Hold endpoint isolates
photo-attach failures so they don't roll back the hold record.
Tests cover: payload shape, bad job_id, hold create with/without photo,
empty qty rejection, empty-signature rejection, sign-off finish, and
the no-milestone-action error path.
Verify on entech: -u fusion_plating_shopfloor --test-tags fp_shopfloor.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Plan task P1.7. Final shared service — standard WO card used on Landing
kanban, Manager Plant Board, and Workflow Funnel. Embeds WorkflowChip,
shows progress bar, priority dot, blocker badge from step.blocker_kind.
Density prop ('compact' vs 'normal') swaps padding for funnel use.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Plan task P1.6. Modal hold-creation form: reason picker, qty split,
optional photo (camera input on mobile), description, mark-for-scrap
toggle. Calls /fp/workspace/hold (added in P1.9). Reason list kept
client-side, keep in sync with fusion.plating.quality.hold.hold_reason.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Plan task P1.5. Modal canvas signature capture using HTML pointer events
+ Odoo Dialog service. Returns image/png dataURI via onSubmit callback;
caller decides what to do with it (e.g. /fp/workspace/sign_off attaches
to fp.job.step).
Canvas stays light even in dark mode for signature legibility.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Plan task P1.4. "Can't start yet — Waiting on Step N: X" block reused
across JobWorkspace step rows and Manager Plant Board cards. Icon set
maps to blocker_kind (predecessor/contract_review/parts_not_received/
racking_required/manager_input). Optional Jump button propagates to
parent via onJump callback.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Plan task P1.3. Bootstraps the tests/ dir and adds the first of 5
shared OWL services. Pill renders fp.job.workflow.state with color
mapping + optional next-action hint.
Per CLAUDE.md "Dark Mode" rule: registered once in web.assets_backend;
Odoo 19 auto-compiles into both bright and dark bundles via the
\$o-webclient-color-scheme SCSS branch.
Version bumped to 19.0.27.0.0 (Phase 1 — Workspace foundation).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Plan task P1.2. Reuses _fp_should_block_predecessors so the new compute
stays in sync with the existing can_start logic. Drives the OWL GateViz
component on the tablet — "Can't start yet — Waiting on Step N: X".
Future work: extend with explicit branches for contract_review /
parts_not_received / racking_required / manager_input as those gate
models mature.
Tests not run locally (no fusion_plating mount in odoo-modsdev).
Verify on entech: -u fusion_plating_jobs --test-tags fp_jobs.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Plan task P1.1. Formats fp.job.name as "WO # <last-segment>" for
tablet/dashboard surfaces. Underlying name field is unchanged so
back-office forms, reports, and emails keep WH/JOB/00001.
Tests not run locally — fusion_plating not mounted in odoo-modsdev
container. Verify on entech: -u fusion_plating_jobs --test-tags fp_jobs.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Aggressive sheet override: flex-basis 100%%, !important on width and
max-width to beat parent flex/media-query constraints. Also overrides
the o_form_sheet_bg wrapper.
Layout at xl (>=1200px) now splits into 3 columns:
- Col 1 (3/12): Your Activities + Bottlenecks
- Col 2 (5/12): ADP Pre + ADP Post + MOD
- Col 3 (4/12): Aging + Other Funders + Recent ADP Exports
Falls back to 5/7 on lg (Col 3 wraps below as full row) and stacked
single column on md and below.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The .o_fc_dashboard .o_form_sheet override wasn't winning specificity
against Odoo's default form-sheet constraints. Added a dedicated class
o_fc_dashboard_sheet directly on the <sheet> element + !important
overrides on max-width, width, and flex to stretch the sheet to the
full container width.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Adds 4 new sections:
- This Month rollup: submitted/approved/delivered/billed counts MTD
- Pipeline $ by stage: pre-submit / submitted / approved / ready-to-bill amounts
- Aging buckets: 30-59d, 60-89d, 90+ days
- Recent ADP Exports: last 5 with totals
Also overrides Odoo's form-sheet max-width on .o_fc_dashboard so the
dashboard uses the full browser width.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The dashboard action existed but no menuitem ever pointed to it (latent
bug in the original module). Adding menu_fusion_claims_dashboard as the
first child of menu_adp_claims_root so the dashboard becomes the default
landing for the Fusion Claims app.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Action-oriented dashboard replacing the existing 4-panel HTML overview:
posting-week banner with live countdown, 3 KPI tiles, 8 funder hotlinks,
ADP + MOD workflow flag tiles, role-aware filtering, dark-mode aware SCSS.
Spec captures all design decisions from the brainstorm session; ready to
hand off to writing-plans.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
User shared their actual published service-rate card. Bundle 9's seeded
numbers were placeholders that no longer match. Realigned the rate card,
added the LIFT & ELEVATING SERVICE class, added the in-shop labour
rate path, added the delivery / pickup charge model, added rush as a
proper tier (distinct from after-hours), and added 30-min increment
rounding on top of the existing 1-hour minimum.
EQUIPMENT CLASS
fusion.repair.product.category gets a new x_fc_equipment_class
selection: 'standard' vs 'lift_elevating'. The published card splits
pricing into two service classes - lift_elevating has higher rates
($160 callout vs $95, $110/h vs $85).
Categories marked lift_elevating in seed:
stairlift, porch_lift, lift_chair (new)
New 'Lift Chair' category seeded (power recliner / lift chair).
CALLOUT RATE CARD
fusion.repair.callout.rate gets:
- equipment_class field (standard / lift_elevating)
- in_shop_labor_rate field (separate $75 vs $85 on-site)
- 'rush' tier value (was missing - rush was implicit via emergency
surcharge from Bundle 8; now a proper tier matching the printed
rate card row 'Rush Service Calls $120')
Re-seeded with the PUBLISHED Westin rate card (exact values):
STANDARD SERVICE
regular $95 callout / $85/h on-site / $75/h in-shop
rush $120 callout / $85/h / $75/h
after_hours $140 callout / $85/h / $75/h
weekend $180 callout / $85/h / $75/h (extension)
holiday $220 callout / $85/h / $75/h (extension)
LIFT & ELEVATING SERVICE
regular $160 callout / $110/h on-site / $110/h in-shop
rush $200 callout / $110/h / $110/h (extension)
after_hours $240 callout / $110/h / $110/h (extension)
weekend $300 callout / $110/h / $110/h (extension)
holiday $360 callout / $110/h / $110/h (extension)
Travel: $0.70 per km, BOTH WAYS, past 25 km, per technician
(matches the per-card '$0.70 per km x 2-way' footnote).
get_for_tier(tier, equipment_class) now resolves with a fallback:
tries (tier, lift_elevating) first, falls back to (tier, standard)
if no lift-specific row exists - so an admin can leave standard rows
as the catch-all and only customise lift for the exceptions.
DELIVERY / PICKUP RATE CARD
New fusion.repair.delivery.charge model + seed of all 7 items from
the printed card:
Local Service Area (within Brampton) ........ $35
Outside Local Area .......................... $60
Rush Pickups / Delivery ..................... $60 + $0.70/km x 2-way
Lift Chair Delivery and Set-Up .............. $120
Hospital Bed Delivery and Set-Up ............ $120
Stairlift Delivery and Set-Up ............... $300
Stairlift Removal ........................... $300
quote_rush(distance_km) helper for the office's delivery scheduling.
New menu: Configuration > Delivery / Pickup Charges.
PRICING ENGINE UPDATES (repair.order._compute_callout_quote)
- Class-aware rate lookup (uses category.equipment_class).
- In-shop mode (x_fc_in_shop=True): skips callout fee + extra-tech +
travel; charges in_shop_labor_rate * hours * techs only. Per the
rate-card footnote 'In-Shop Labour Rate'.
- 30-min increment rounding ON TOP of the 1-hour floor:
billable_h = max(ceil(actual * 2) / 2, min_hours)
-> 20-min work bills 1.0 h
-> 75-min work bills 1.5 h
-> 95-min work bills 2.0 h
- Improved breakdown text shows the rate-card row name + class +
pro-ration math so the client can see how the total was computed.
NEW FIELDS
repair.order:
x_fc_in_shop (Boolean) - flip to switch the quote engine to
in-shop mode.
x_fc_callout_tier now includes 'rush' as a value (was missing).
visit-report wizard:
callout_in_shop related field - tech can flip the mode on-site if
the work was actually done in-store after pickup.
MIGRATION SCRIPT
migrations/19.0.2.1.0/post-migration.py runs once on existing
installs:
1. Updates stairlift / porch_lift / lift_chair categories
equipment_class -> lift_elevating
2. Wipes the 4 Bundle 9 rate-card xml_ids so the new noupdate=1
seed creates them with the correct printed values.
Fresh installs get the right values directly from the seed XML.
Admin-created custom rate rows (no xml_id) are NEVER touched.
VERIFIED END-TO-END (0 bugs across 28 checks)
Rate card matches printed values exactly:
regular/standard = $95/$85h/$75h PASS
rush/standard = $120/$85h/$75h PASS
after_hours/standard = $140/$85h/$75h PASS
regular/lift = $160/$110h/$110h PASS
Six end-to-end quote scenarios:
A. Standard 12km 20-min -> $180 ($95 + 1h*$85)
B. Lift 12km 20-min -> $270 ($160 + 1h*$110)
C. Rush 30km 1.2h -> $254.50
($120 + ceil(2.4)/2=1.5h * $85 + 5km*2*$0.70 = $7)
D. After-hours lift 2-tech 35km 2.6h -> $928.00
($240 + ceil(5.2)/2=3.0h * $110 * 2 + 10km*2*$0.70*2)
E. In-shop standard 2h -> $150 (2h * $75 in-shop, no callout)
F. In-shop lift 1.5h -> $165 (1.5h * $110 in-shop)
Seven delivery rates loaded with correct amounts; rush 40km calc
= $81 ($60 base + 15km*2*$0.70).
Stairlift / Porch Lift / Lift Chair categories correctly marked
lift_elevating; rest stay standard.
Bumped to 19.0.2.1.0.
Co-authored-by: Cursor <cursoragent@cursor.com>
New module `fusion_service_charges` that creates the standard
service-billing product catalog for Westin Healthcare and Mobility
Specialties:
Standard Service
SVC-STD-CALL Service Call (incl. 30 min) $95
SVC-STD-LABOUR Standard Labour (hourly) $85
SVC-INSHOP-LABOUR In-Shop Labour (hourly) $75
SVC-RUSH-CALL Rush Service Call $120
SVC-AH-CALL After-Hours Service Call $140
Lift & Elevating
SVC-LIFT-CALL Lift Service Call (incl. 30 min) $160
SVC-LIFT-LABOUR Lift Labour (hourly) $110
Delivery / Pickup
DEL-LOCAL Local (within Brampton) $35
DEL-OUT Outside Local Area $60
DEL-RUSH Rush Delivery / Pickup $60
DEL-LIFT-CHAIR Lift Chair Delivery + Set-up $120
DEL-HOSP-BED Hospital Bed Delivery + Set-up $120
DEL-STAIRLIFT Stairlift Delivery + Set-up $300
SVC-STAIRLIFT-RM Stairlift Removal $300
Loading pattern (intentional):
- Products created via post_init_hook on FIRST install only.
- Manifest's `data` list is EMPTY so no XML is loaded on `-u`.
- Hook is idempotent — sentinel ir.model.data xmlid check skips
records that already exist. Safe to re-run.
- User edits / deletes survive every upgrade (proven on entech-
westin: edited SVC-STD-CALL price to $999.99 → ran -u → price
stuck. Reset to $95 after test.).
- Uninstall + reinstall does re-seed (ir.model.data sentinels drop
on uninstall, fresh install treats it as new).
Per-km surcharges (Rush, Outside Local, After-Hours) are noted in
the product description so the dispatcher knows to add a separate
mileage line. Formula-based pricelist for auto-mileage is out of
scope — matches current manual workflow on both shops.
Odoo 19 compatibility: dropped uom_po_id from the create vals
(retired in 18; uom_id is now the single source of truth for sale
and purchase UoM on product.template).
Deployed and verified on:
- odoo-westin / westin-v19 (Docker: odoo-dev-app) — 14 products
- odoo-mobility / mobility (Docker: odoo-mobility-app) — 14 products
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Full home-service pricing engine plus the store labor warranty model. The
call price now itemises base callout + extra techs + hourly labour (with
the 30-min-included + 1-hour-minimum rule) + travel both ways past
threshold, with three independent waive paths: in-warranty / manager
override / sales-rep override. CS cannot waive (RBAC).
NEW MODELS
fusion.repair.callout.rate (rate card)
Per (tier, company) row. Tiers: regular / after_hours / weekend / holiday.
Fields:
- base_callout_fee (INCLUDES first 30 min for inspection / report)
- second_tech_fee + additional_tech_fee (3rd, 4th tech)
- hourly_labor_rate + minimum_labor_hours (default 1.0 floor)
- travel_distance_threshold_km + travel_per_km_fee
- effective_from (newer rows supersede older)
Seeded with 4 default rows (regular $120/$95/0.85, after-hours
$180/$140/1.10, weekend $240/$170/1.35, holiday $300/$200/1.50).
fusion.repair.labor.warranty (store labor warranty)
Per (partner, product/lot, sale_order) record with warranty_years +
start_date + computed end_date. State machine: active / expired / void
/ consumed. Void reasons spec'd by the user: user_negligence /
gross_negligence / misuse / over_recommended_use / accidental_damage
/ not_covered_part / other.
find_active_for(partner, product, lot) - lot-first then product+partner
then partner-only fallback so warranty resolution survives partner-
contact / product-variant differences.
action_void(reason, notes) - manager-only; audit stamps voided_by_id
+ voided_at + reason; posts chatter.
PRODUCT EXTENSION
product.template.x_fc_labor_warranty_years (Integer, default 0).
SALE-ORDER EXTENSION
sale.order.action_confirm now also runs _fc_spawn_labor_warranties()
which creates one fusion.repair.labor.warranty per unit of any product
with x_fc_labor_warranty_years > 0. Lives alongside the existing
service-plan spawn so a 5y-LW stairlift sold with a maintenance plan
spawns both records in one go.
PRICING ENGINE ON REPAIR.ORDER
9 new fields:
x_fc_callout_tier (regular/after_hours/weekend/holiday)
x_fc_callout_distance_km (one-way; system bills both ways)
x_fc_callout_techs (1, 2, 3+)
x_fc_callout_labor_hours (hours above the 30 min the callout covers)
x_fc_labor_warranty_id (auto-resolved on visit)
x_fc_labor_warranty_status (not_checked / eligible / not_covered /
expired / void_misuse / waived)
x_fc_labor_waived + _by_id + _at + _reason
6 computed quote fields:
x_fc_quote_callout_base (base_callout_fee)
x_fc_quote_extra_techs (second + additional fees)
x_fc_quote_labor (max(hours, min_hours) * rate * techs)
x_fc_quote_travel (max(distance - threshold, 0) * 2 * per_km * techs)
x_fc_quote_waived (= labor if warranty eligible OR labor waived)
x_fc_quote_total (sum minus waived; stored, indexable)
+ a human-readable x_fc_quote_breakdown_text used in the email template.
3 new actions:
action_check_labor_warranty (anyone) - resolves the warranty and
stamps x_fc_labor_warranty_status. Called automatically by the
visit-report wizard.
action_waive_labor_fee (SECURITY GATED) - raises UserError unless
caller is in group_fusion_repairs_manager OR
group_fusion_repairs_sales_rep. CS users get the explicit message
'Only Repairs Managers and Sales Reps can waive the labor fee.'
action_acknowledge_rush - Bundle 8 carryover.
SECURITY
New group_fusion_repairs_sales_rep
Independent group so a sales rep can waive labor on their accounts
without becoming a Repairs Dispatcher / Manager. Manager IMPLIES
sales_rep so managers automatically inherit the right.
ACLs: callout.rate user-read / manager-full; labor.warranty user-read /
sales_rep-write / manager-full / technician-read+write.
VISIT-REPORT WIZARD EXTENSIONS
Pricing block (visible when outcome=completed):
callout_tier / techs / distance_km / labor_hours_used (default 1.0
minimum). Live quote_total_preview + breakdown shown to the tech so
they can confirm the price with the client right at the door.
Warranty block:
labor_warranty_id_preview + labor_warranty_status_preview (badge
coloured by status). 'warranty_void_reason' selection lets the tech
void the warranty in real time when they find misuse / negligence /
accidental damage - on submit the matching warranty record is voided
permanently (action_void) AND the repair's labor charge re-computes
without the waive.
On confirm the wizard:
1. Persists callout_labor_hours_used to the repair
2. Calls repair.action_check_labor_warranty()
3. If warranty_void_reason set + warranty resolved -> voids it,
posts chatter, repair labor_warranty_status -> void_misuse
NAVIGATION
Repair form 4 new header buttons:
Check Labor Warranty (anyone)
Waive Labor Fee (sales_rep + manager only, server-side gated)
(plus the Bundle 8 Squeeze + Ack Rush from before)
New 'Callout Pricing' notebook tab on repair form with:
inputs, warranty/waiver, and the 6-line quote breakdown.
New menus:
Fusion Repairs > Labor Warranties
Configuration > Callout Rate Card
Configuration > Emergency Surcharges (Bundle 8 carryover)
VERIFICATION END-TO-END (7 scenarios, 0 bugs)
A. Sale of a product with 5y LW -> LW-00002 spawned, expires 2031-05-21.
B. In-warranty regular 12km 20-min repair:
base 120 + labor 95 - waived 95 = $120 (callout only)
C. After-hours 2-tech 40km 1.5h, NO warranty:
180 + 90 + (1.5*140*2) + (15*2*1.10*2) = $756.00 exact
D. In-warranty visit -> tech ticks misuse void_reason:
Warranty record -> state=void / reason=misuse.
Repair labor_warranty_status -> void_misuse.
Quote re-computes WITHOUT waive: labor 1.5 * 95 = $142.50 charged.
E. Manager waives labor on a no-warranty repair:
Pre-waive $310 -> post-waive $120 (labor $190 -> waived).
Audit: waived_by_id stamped to gsingh@.
F. CS rep tries to waive: correctly denied with the spec'd error
'Only Repairs Managers and Sales Reps can waive the labor fee.'
G. Weekend 1-tech 30km 30-min:
240 + (1.0*170) + (5*2*1.35) = $423.50 exact (min-1h floor
correctly applied to the 0.5h actual work).
Bumped to 19.0.2.0.0 (minor version bump - new public-facing model).
Co-authored-by: Cursor <cursoragent@cursor.com>
The grumpy-old-customer-with-broken-stairlift scenario. Four real workflows
the office faces every week, with comms baked in so the client never has to
call back asking for status.
NEW MODELS
- fusion.repair.emergency.charge (rate card)
Per (category, tier) rate with per_tech_multiplier; 5 tiers
(same_day / next_day / after_hours / weekend / holiday). Each category
can have its own rates - bed motors need 2 techs, stairlift is single.
Seeded with realistic Westin rates: stairlift same-day $250, weekend
$450; porch lift same-day $300; bed same-day $175 with 0.6 multiplier
(2-tech jobs frequent); powerchair same-day $200.
- fusion.repair.part.order (procurement-facing record)
One per distinct part the tech needs from the manufacturer. Carries
description + OEM # + manufacturer + quantity + photos + notes.
4-state lifecycle: draft -> ordered -> received -> fitted (or
cancelled). On state transitions:
draft -> ordered: email client "ordered, expected by X"
ordered -> received: email client "arrived, scheduling return visit"
+ auto-create follow-up dispatch task when ALL
outstanding parts on the repair have arrived.
REPAIR.ORDER EXTENSIONS
- Rush fields: x_fc_rush_requested, x_fc_rush_tier,
x_fc_rush_techs_required, x_fc_rush_surcharge (computed via rate card),
x_fc_rush_acknowledged_at + x_fc_rush_acknowledged_by_id (audit trail
proving CS got verbal OK before charging).
- Parts-awaiting fields: x_fc_parts_awaiting + x_fc_parts_eta_date +
x_fc_part_order_ids One2many + x_fc_part_order_count.
- New methods:
* action_acknowledge_rush() - one-click "client agreed" with audit.
* action_squeeze_into_today() - picks the lightest-loaded skilled tech,
finds their first free 1-hour slot between 9am-6pm, schedules the
task in it, sends:
1) live bus.bus push to the tech (sticky notification in their
web client - so they see it MID-SHIFT)
2) rush-alert email (force_send=True - this can't wait in the queue)
3) chatter post on the tech task itself
Validates against fusion_tasks' time-conflict rule by passing
force_schedule via context (intake.service honours it).
* action_view_part_orders() - smart button.
WIZARD EXTENSIONS
- repair.intake.wizard:
New rush_requested + rush_tier + rush_techs_required + rush_acknowledged
controls. Live rush_surcharge_preview compute shows CS the price in
real-time as they change category / tier / tech count. Yellow alert
reminds CS to read the price to the client BEFORE submitting.
- repair.visit.report.wizard:
New outcome radio: completed / parts_needed / rescheduled.
When outcome=parts_needed, needs_parts_line_ids One2many appears for
the tech to capture each part (description, OEM, manufacturer, qty,
lead days, notes, photos). On submit each line creates a
fusion.repair.part.order, the repair flips to x_fc_parts_awaiting=True
with an ETA, and the client gets the "we found the problem, here's the
plan" email immediately.
INTAKE SERVICE
- _create_dispatch_task now honours force_schedule (date + time_start +
time_end) via context so squeeze + auto-redispatch don't crash on
fusion_tasks' time-window validator.
- _create_single_repair carries rush_requested/tier/techs through to
the new repair fields.
MAIL TEMPLATES (4 new)
- email_template_rush_tech_alert: red 4px accent, address + phone + the
$surcharge - what the tech needs to know mid-shift.
- email_template_repair_awaiting_parts: amber accent, "we found the
problem, parts ordered, return visit ~ETA, no action needed".
- email_template_parts_ordered: blue, per-part confirmation.
- email_template_parts_received: green, "arrived, office will call to
confirm visit".
UI / NAVIGATION
- Backend wizard: rush controls + live surcharge preview + verbal-OK alert.
- repair.order form: new Rush / Parts notebook tab with all the fields
+ linked part orders list. Two new header buttons (Squeeze into
Today / Client Agreed to Rush Price). Two new search filters
(Rush, Awaiting Parts).
- Part Order form: statusbar with the 4 transitions + Cancel; notes +
photos notebook tabs; full chatter for audit.
- Menus: 'Parts to Order' under root; 'Emergency Surcharges' under
Configuration.
SECURITY
- 8 new ACL entries (emergency_charge user/manager; part_order
user/dispatcher/manager/technician; visit_report partline for office
and field tech). Office sees parts but only managers can edit
emergency rates.
Verified end-to-end on local westin-v19 - all 4 scenarios green:
S1 Same-day rush stairlift -> $250 surcharge, ack stamped, squeeze
assigned garry@ at first free 1h slot today, alert email queued,
chatter posted.
S2 Next-day priority bed -> $0 surcharge (no rate seeded for bed
next_day - office can configure), 4 emails queued (client + office).
S3 2-tech weekend stairlift -> $675 (450 base + 0.5x base for 2nd tech).
S4 Parts-needed visit-report -> 2 PART-#### records created, repair
awaiting_parts=True, ETA=2026-06-06, office activity scheduled,
client email sent. Marking part ordered -> client mail. Marking
all parts received -> auto-dispatch follow-up + client mail.
Bumped to 19.0.1.9.1.
Co-authored-by: Cursor <cursoragent@cursor.com>
Full end-to-end walk acting as customer, CS rep, dispatcher, technician,
and manager surfaced 6 real bugs (1 critical state-machine, 4 missing UX
wires, 1 docstring). Server endpoints existed for everything but several
were not wired into the templates.
B1 (HIGH) - Visit-report wizard never closed the repair
Tech submitted visit -> state stayed 'draft' -> x_fc_done_at never
stamped -> NPS cron never fired -> the whole post-visit flow died
silently. Customers never got their NPS email.
Fix: action_confirm() now drives the Odoo native state machine
draft -> action_validate (with _action_repair_confirm fallback) ->
action_repair_start -> action_repair_end. Each step guarded by the
current state and exception-logged. Leaves the repair open if:
- requires_requote=True (variance flag - office must re-quote)
- no_show=True (office reschedules)
- x_fc_is_quote_only (still a quote)
- found_another_issue spawned a stub
Posts a clear chatter line on success or failure.
Verified: e2e walk now shows state=done + x_fc_done_at stamped +
NPS cron fires + flags x_fc_nps_email_sent=True.
B2 (HIGH) - /repair/new form never called /repair/self_check
The AI self-check engine was the headline weekend feature but it was
invisible to the client. The endpoint worked server-side, just had
no frontend.
Fix: new portal_client_repair.js (Interaction class, registered on
registry.category('public.interactions')). 'Try 1-3 safe self-check
steps first' button POSTs to /repair/self_check, renders steps via
createElement + textContent (no innerHTML - all server output is
treated as untrusted text). Shows the AI's safety disclaimer on
every result. On escalate_immediately, shows a clear 'submit the
form, we'll come to you' message instead of the steps.
Verified: HTTP POST returns full JSON with instruction +
expected_result + disclaimer; new button + result panel appear in
rendered HTML.
B3 (HIGH) - No phone-lookup UI for returning clients
Same problem - endpoint existed but no UI. Returning clients had to
retype everything from scratch.
Fix:
- lookup_phone now returns a 'partners' array (id, name, email,
street, city) - cap of 3 results, rate-limited, every match logged
at INFO level for audit. Privacy compromise: a phone holder
deserves to see their own pre-fill; rate limit caps harvesting.
- JS lookup widget at the top of the form posts to /repair/lookup_phone
and pre-fills the 5 contact fields + writes the partner_id to a
hidden #fr_known_partner_id input.
- controller /repair/submit now trusts known_partner_id if present
(skips the phone re-match) so we don't create duplicate partners
when the lookup widget already identified the right one.
Verified: HTTP POST returns the 2 partner records we have for
+19055551234 with full id/name/email/street/city.
B4 (MEDIUM) - /repair?sn=<serial> from QR sticker did nothing
Spec: 'Client scans QR sticker - portal pre-fills the unit info.'
Reality: the form had no serial field; ?sn= was ignored.
Fix: new _resolve_serial_info(serial) on the controller resolves
the lot via stock.lot.search([('name','=',sn)]) and returns
{serial, lot_id, product_id, product_name, category_id}. Both
/repair (landing) and /repair/new pass it as serial_info template
context. Templates show 'Recognized X (Serial: Y)' + auto-select
the matching category in the dropdown. Hidden #fr_serial_number
carries it through to /repair/submit, which attaches the lot_id +
uses the QR category as fallback if user didn't pick one.
Verified: ?sn=stella23-20040164 produces 'Pre-filled from QR scan:'
banner + hidden input populated.
B5 (MEDIUM) - No upsell after submit
Spec required an upsell - 'reduce future calls'. Page was a bare
'Got it'.
Fix: /repair/thanks now shows a 2-card layout:
- 'Want to avoid this next time?' with 4 bullets (priority booking,
free inspection cert, discounted parts, annual reminder) +
'See our maintenance plans' CTA to /shop?category=maintenance
- 'What happens next' 4-step bulleted explanation
Verified: both cards render.
B6 (LOW) - SyntaxWarning '\-->' in repair_service_plan.py
Made the module docstring a raw string (r''') so the ASCII flowchart
arrows don't trigger Python's invalid-escape-sequence warning.
Bumped to 19.0.1.8.0.
Co-authored-by: Cursor <cursoragent@cursor.com>
T3 Labour timer on technician task
- Two new fields on fusion.technician.task: x_fc_timer_running_since
(Datetime) + x_fc_timer_accumulated_minutes (Float).
- action_timer_start / action_timer_stop methods, idempotent (start when
already running is a no-op, stop when not running is a no-op).
- Multiple start/stop cycles accumulate into the same total.
- Two header buttons (Start Timer green / Stop Timer amber), invisible
based on the running_since field so the right one shows at any time.
- Stop posts a chatter line 'Labour timer stopped. Added X.X min, total
Y.Y min.' so audit history shows every shift.
T4 Client signature on visit report
- New client_signature Binary field on the visit-report wizard with
Odoo native widget='signature' that draws on canvas + base64-encodes
the PNG.
- client_signature_name Char for typed name (audit).
- Persisted as an ir.attachment on the repair.order via the new
_persist_mobile_artefacts helper.
- Chatter post 'Client signature captured (Jane Smith).'.
T6 Replaced parts - serial capture
- parts_serial_capture Text on the wizard (one per line per the spec).
- On confirm, posted to chatter wrapped in <pre> so line breaks survive.
- Used by OEM warranty filing in future M8.
T7 Client no-show photo proof
- no_show Boolean + no_show_photo Binary with widget='image' (visible
only when no_show=True via Odoo 19 invisible= conditional).
- Photo saved as ir.attachment on the repair when present.
- Chatter post 'Visit recorded as client no-show (photo attached)'.
Verified end-to-end on local westin-v19:
T3 timer started -> 2s sleep -> stopped -> 0.0357 min recorded
T4 attachment 'signature-RO-202605-17.png' created on repair
T6 chatter shows 'SN-AAA-111 / SN-BBB-222'
T4 chatter shows 'Client signature captured (Jane Smith)'
Bumped to 19.0.1.7.0.
Co-authored-by: Cursor <cursoragent@cursor.com>
M9 margin per repair
- New non-stored computes on repair.order: x_fc_revenue, x_fc_labour_cost,
x_fc_parts_cost, x_fc_margin, x_fc_margin_pct.
- Revenue: sum of posted out_invoice.amount_untaxed on the repair's sale
order (handles partial / multi invoice scenarios).
- Labour: sum of (task.duration_hours x technician.x_fc_tech_cost_rate)
over COMPLETED visits only - avoids counting scheduled-but-not-done time.
- Parts: sum of standard_price x qty for stock moves where
repair_line_type='add' (parts consumed, not removed).
- New 'Margin' notebook tab on repair.order form, manager-group gated.
M7 failure analytics on the dashboard
- Three new keys in get_dashboard_data():
* failures_by_product - top 8 products by repair_count in last 90 days
via _read_group (efficient - no record load)
* failures_by_symptom - top 8 x_fc_issue_category values
* margin_summary - revenue/labour/parts/margin/margin_pct + sample_size
over the same 90-day window
- Three new tiles on the OWL dashboard 'Last 90 Days' section:
Margin Summary (revenue/labour/parts/margin breakdown),
Failure Rate by Product, Failure Rate by Symptom.
- New formatMoney + formatPercent helpers on the dashboard JS so values
display as 'CAD 12,345' rather than raw floats.
Verified end-to-end on local westin-v19:
Dashboard returned all 9 expected keys.
Top product: 'M6 X 27 THREADED BARREL' (2 repairs) - actual test data.
Margin summary over 26 repairs (dev has $0 invoices so values 0.0,
but the compute path is exercised and shapes are correct).
Bumped to 19.0.1.6.0.
Co-authored-by: Cursor <cursoragent@cursor.com>
New models
- fusion.repair.service.plan.subscription
Tracks pre-paid maintenance packages: partner, plan product, optional
category restriction, visits_included / visits_used / visits_remaining,
start_date / end_date, computed state (active/exhausted/expired/cancelled),
burn_history One2many. PLAN-NNNNN sequence.
- fusion.repair.service.plan.burn
One row per maintenance visit that consumed a plan visit - feeds the
Burn History tab on the subscription form.
product.template extensions
- x_fc_is_service_plan boolean toggle
- x_fc_plan_visits_included (default 4)
- x_fc_plan_duration_months (default 12)
- x_fc_plan_category_id - if set, only burns for repairs in that category
(e.g. an Annual Stairlift Maintenance plan does not burn for wheelchair
repairs)
sale.order.action_confirm() override
- For each order line whose product has x_fc_is_service_plan=True,
spawns one fusion.repair.service.plan.subscription per qty unit.
- Start date = today; end date = today + plan_duration_months
(relativedelta - correct month boundaries).
Visit report wizard
- New _burn_service_plan_visit(repair) call from action_confirm() finds
the matching active subscription and burns one visit + posts a chatter
note "Visit burned for repair X. N of M remaining." on the subscription.
- Skips quote-only repairs.
- The wizard does NOT zero out the invoice - the burn is informational;
the office reconciles plan credits in their accounting workflow.
Backend
- Service Plans menu under Fusion Repairs root.
- List view colour-coded by state.
- Form with statusbar + cancel button + Burn History notebook.
- Service Plan tab added to product.template form (manager only).
- ACL: User read; Dispatcher write/create; Manager full + unlink.
Verified end-to-end on local westin-v19:
Created plan product 'Annual Stairlift Maintenance - 4 Visits'
Sold it via sale.order -> PLAN-00001 auto-created
(visits_included=4, end_date=2027-05-21)
Submitted visit-report on a stairlift repair -> visits_used=1
remaining=3 (correctly category-matched).
Bumped to 19.0.1.5.0.
Co-authored-by: Cursor <cursoragent@cursor.com>
H1+H2: Field technicians had perm_create=1 perm_write=1 on inspection
certs (could forge or edit issued certs). Reduced to read-only - the
visit-report wizard already sudos when creating new certs from a tech
visit. Added rule_inspection_cert_readonly for the dispatcher group so
even dispatchers cannot edit already-issued certs; only the manager can
revoke/correct. Sealed audit trail.
H3: Replaced display:flex / gap (which wkhtmltopdf 0.12 renders as a
vertical stack) with inline-block + margin in the certificate PDF.
Footer uses float left/right for the cert-number / inspector signature
line so the layout survives wkhtmltopdf rendering.
Bumped to 19.0.1.4.1.
Co-authored-by: Cursor <cursoragent@cursor.com>
New fusion.repair.inspection.certificate model for the annual safety
inspections required on stairlifts, porch lifts, and power wheelchairs
in many jurisdictions.
Model
- mail.thread chatter-tracked; fields: name (CERT-YYYY-NNNN auto-seq),
partner_id, product_id (filtered to safety-critical categories), lot_id,
repair_order_id back-link, inspector_user_id (must be field staff),
jurisdiction (selection: Ontario / BC / Alberta / Quebec / Other),
issued_date, valid_for_months (default 12), expiry_date (computed,
stored, uses relativedelta - correct month boundaries), status
(non-stored compute: valid / expiring / expired / revoked), revoked,
notes, last_reminder_band.
- Unique constraint on certificate number (models.Constraint, not
_sql_constraints, per project rule).
- Sequence 'fusion.repair.inspection.certificate' with use_date_range=True
so the counter resets each year (CERT-2026-0001 ... CERT-2027-0001).
Visit report integration
- New issue_inspection_cert checkbox on fusion.repair.visit.report.wizard.
- When ticked AND the repair's category is safety_critical, action_confirm()
creates the certificate via _create_inspection_certificate() and
redirects to the cert form so the tech can print immediately.
- Non-safety-critical equipment quietly skips with a chatter note
explaining why.
PDF report
- web.html_container + web.external_layout, model bound so it appears
as a Print action on the certificate form.
- 'Certificate of Inspection' / 'Safety Inspected' gold-banner layout
with client name, equipment, serial, jurisdiction, issued + expiry
dates, inspector signature line, and the certificate number.
- Print Certificate button in form header.
Daily cron
- cron_send_expiry_reminders runs at 09:00, sends two band-tracked
reminders (30 days + 7 days before expiry) to the client.
- New mail.template email_template_inspection_expiry_reminder with
4px amber accent, certificate ref, equipment, expiry date, and a
CTA to call to book the re-inspection visit.
- last_reminder_band on the cert prevents re-sending the same band.
Backend wiring
- New menu entry 'Fusion Repairs > Inspection Certificates'.
- ACL: User read, Dispatcher write, Manager unlink. Field technicians
can create (they need to issue from the field).
- List view with red/amber/green status decoration.
- Form with statusbar, header buttons (Print, Revoke with confirm),
chatter.
Verified end-to-end on local westin-v19:
Stairlift repair RO-202605-15 -> visit-report with issue_inspection_cert=True
-> CERT-2026-0001 issued (status=valid, expires 2027-05-21)
Cert CERT-2026-0002 expiring in 30 days -> cron flagged
last_reminder_band='30' (would email client).
Bumped to 19.0.1.4.0 (minor bump for the new public-facing capability).
Co-authored-by: Cursor <cursoragent@cursor.com>
HIGH
H1 X2 reminder flag was per-repair - multi-visit repairs missed reminders
Moved x_fc_day_before_reminder_sent off repair.order onto
fusion.technician.task so each scheduled visit is tracked separately.
Cron now walks tasks directly with state-narrowed repair filter
(confirmed/under_repair only, drops L1's draft inclusion).
H2 X4 NPS cron used write_date - moved on every chatter/invoice write
Added x_fc_done_at Datetime on repair.order, stamped on the first
transition to state=done via write() override. Cron filters on
('x_fc_done_at', '<=', cutoff) instead of write_date.
H3 X2 template's [:1] slice picked an arbitrary task, not tomorrow's
Cron now passes the specific task via with_context(reminder_task_id=...).
Template fetches that task by id; falls back to [:1] only for manual
sends so chatter Send Email composer still works.
H4 NPS Google-Search fallback URL not URL-encoded - breaks on &/spaces
Template now uses url_encode({'q': company_name}) so "Westin & Sons"
produces a working URL instead of truncating at the ampersand.
H5 + L1 Loaner cron fired on drafts and used create_date instead of schedule_date
Domain rewritten to: state in ('confirmed','under_repair'), exclude
quote-only repairs, and EITHER schedule_date <= cutoff OR (schedule_date
is False AND create_date <= cutoff). Added limit=200 ordered by
create_date desc (M6).
MEDIUM
M1 Function-level datetime imports moved to module top
date, datetime, timedelta imported once at the top of repair_order.py,
removed from cron_send_day_before_reminders, cron_send_post_visit_nps,
cron_offer_loaner_for_long_repairs.
M2 _notifications_enabled duplicated - promoted to single source
repair_order._notifications_enabled now delegates to
fusion.repair.intake.service._notifications_enabled() (with a fallback
ICP read if the service AbstractModel isn't available).
M3 self.env.get('model') -> 'model' in self.env (Odoo standard idiom)
Two call sites in repair_order.py converted.
M4 + M5 Bare 'except: continue' + missing logger - operational blindness
Added import logging + _logger to repair_order.py. All three crons now
log exceptions with _logger.exception(). Activity-type ref check now
warns + returns early if the xml id is missing (instead of passing
activity_type_id=False which raises). For X2 and X4 the flag is set
regardless of send-success so we don't retry indefinitely on
permanently-misconfigured partners.
M6 Loaner cron has limit=200 + order='create_date desc'
Caps blast radius if 5000 stale draft repairs ever accumulate.
L1 X2 state filter tightened: was ('not in', ('done','cancel')), now
('in', ('confirmed','under_repair')) so drafts and quote-only don't
email "your tech is coming tomorrow".
Verified - upgrade clean, no errors. Bumped to 19.0.1.3.1.
Co-authored-by: Cursor <cursoragent@cursor.com>
X2 Day-before visit reminder email
- New cron 'Fusion Repairs: Day-before visit reminders' (daily at 08:00)
walks repair.order records with at least one linked
fusion.technician.task scheduled for tomorrow and not yet reminded.
- Sends mail.template email_template_visit_day_before to the client.
- New x_fc_day_before_reminder_sent flag (copy=False) so the cron
never re-sends the same reminder.
- Template uses 4px blue accent, 600px max-width, shows the scheduled
date + technician name + equipment, with a 'reply to reschedule' note.
- Verified: cron flagged the test repair x_fc_day_before_reminder_sent=True
after running.
X4 Post-visit NPS / Google review email
- New cron 'Fusion Repairs: Send post-visit NPS emails' (hourly)
finds repairs in state='done' with write_date >= 24h ago and no NPS
email sent. Sends mail.template email_template_post_visit_nps.
- New x_fc_nps_email_sent flag so we never re-pester clients.
- Template uses 4px green accent + 'Leave a Google review' CTA button
linking to res.company.x_fc_google_review_url (or a sensible Google
search fallback when the company hasn't configured a review URL).
M3 Loaner auto-offer for long-running repairs
- Soft-bridges fusion_loaners_management without a hard dep -
cron_offer_loaner_for_long_repairs returns immediately if the
fusion.loaner.checkout model isn't installed.
- Walks repair.order records open longer than
fusion_repairs.loaner_offer_threshold_days (ICP, default 3 days)
with no existing loaner-offer activity.
- Posts a 'Repair: Offer Loaner' activity (new mail.activity.type)
assigned to the repair responsible.
- New x_fc_loaner_offered flag to prevent daily re-posting.
- Manual 'Offer Loaner' button on repair header opens the
fusion.loaner.checkout wizard pre-filled with partner + SO.
- Daily cron runs at 08:30.
Email + ICP + cron wiring:
- 2 new mail.template records (visit_day_before, post_visit_nps)
- 1 new mail.activity.type (loaner_offer)
- 3 new ir.cron records (day-before, NPS, loaner)
- 1 new ir.config_parameter (loaner_offer_threshold_days)
- 1 new header button (Offer Loaner) on repair.order
Verified end-to-end on local westin-v19:
X2 setup repair: RO-202605-12 task: TASK-00045
day-before flag after cron: True (expected True)
M3 loaner model not installed - cron correctly no-op'd
(no flag set, no activity posted, no error - the soft-dep guard works)
Bumped to 19.0.1.3.0.
Co-authored-by: Cursor <cursoragent@cursor.com>
CRITICAL
C1 Cron re-pages same on-call user forever
page_on_call() now excludes the currently paged user (not just
acknowledged users) so the 15-min escalation cron actually moves
to the next priority. Removed the dead `already` var in the cron.
Verified: page 1 -> gsingh@..., page 2 -> ak@... (different user).
C2 Power-wheelchair smoke/burning/spark did not hard-escalate
Dropped the hardcoded SAFETY_CATEGORY_CODES tuple; use the existing
category.safety_critical Boolean instead. Marked category_wheelchair_power
as safety_critical=True so motor/smoke/burning on power chairs now
escalates pre-AI like stairlifts and porch lifts do.
Verified: powerchair + smoke -> escalate=True.
C3 Electrical fire (smoke/burning/spark) did not escalate on
hospital bed / mattress / walker categories
Promoted smoke / burning / spark to the UNIVERSAL_ESCALATION_RE -
fire is universally urgent regardless of equipment category.
Verified: hospital bed + "motor smells like burning" -> escalate=True.
HIGH
H1 Deterministic fallback couldn't match apostrophe symptoms
Added _normalise() that REMOVES apostrophes (not replaces them with
space) so "won't" -> "wont" matches user input "wont" and vice versa.
Handles straight, curly, and modifier-letter apostrophes.
Verified: "bed wont move" -> matches the "won't move" rule (1 step).
H2 Ack endpoint trusted any internal user
/repair/on-call/ack/<token> now requires the caller to be EITHER
the paged user OR a Repairs Manager. Denied attempts render the
invalid-token page and log a warning.
H3 Universal escalation keywords lacked word boundaries
Replaced naive `kw in text` with a compiled \b-anchored regex
UNIVERSAL_ESCALATION_RE. Likewise SAFETY_SYMPTOMS_RE for category-
scoped symptoms with won.?t to handle the apostrophe variant.
"unhurt" no longer matches "hurt", "firearm" no longer matches "fire".
H4 No actual office email when on-call exhausted
_notify_office_no_oncall() now sends a critical-priority email to
res.company.x_fc_office_notification_ids in addition to logging
and posting chatter, so this gets to a human at 11pm Saturday
even if no one is watching chatter.
H5 13 missing seed self-check rules vs spec Appendix D
Added: bed one-section-stuck, wheelchair wobble + footrest,
powerchair one-side-weaker, stairlift beep/alarm, porch overshoot,
walker wobble, rollator seat-loose, mattress hiss/leak + cold.
10 added (27 total) - within rounding distance of the spec's "30".
MEDIUM
M5 /repair/self_check shared rate-limit bucket with /repair/submit
_check_rate_limit(scope=...) - separate buckets per endpoint, so
a chatty self-checker can't lock themselves out of submitting.
Per-scope ICP cap key (fusion_repairs.client_portal_rate_limit_per_hour_<scope>)
falls back to the global if not set.
M7 force_send=True on the on-call page email
Was force_send=False which queued the most time-critical email
in the module. Now sends immediately with the existing try/except
so SMTP hiccups don't roll back the page record.
M8 QR generation swallowed all errors silently
_logger.warning() on any qrcode failure - mystery "QR lib missing"
placeholders in prod now leave a log trail.
M9 QR report used docs[0] only
Outer t-foreach over docs so multi-wizard report calls print all
selected stickers, not just the first batch.
M10 + M11
- Added models.Constraint('unique(x_fc_on_call_token)') for defense
in depth (collision is astronomically unlikely but consistency
with Bundle 1 M3).
- _send_page_email() returns True/False; _post_chatter only fires
on success. On failure a different chatter line says "page email
failed - verify SMTP".
LOW
L6 find_next_on_call() now filters by company_ids (cross-company safe).
Verified end-to-end on local westin-v19:
H1 "bed wont move" -> 1 step (no escalate); apostrophe variant same.
C1 page 1 -> gsingh; page 2 -> ak (different).
C2 powerchair+smoke -> escalate=True.
C3 bed+burning -> escalate=True.
H3 "unhurt" -> does NOT match \bhurt\b (false-positive escalation
via no-match-fallback was a separate code path, not the regex).
Bumped to 19.0.1.2.2.
Co-authored-by: Cursor <cursoragent@cursor.com>
CL6/CL7 AI self-check engine
- New fusion.repair.ai.service AbstractModel with single guardrailed
suggest_self_check(category_id, symptoms, urgency) entry point.
- Hard-escalation FIRST (before any AI call): stairlift / porch lift +
safety symptoms (smoke / burning / spark / stuck / motor), OR any
mention of fire / injury / hurt / bleeding / trapped, OR urgency=safety
-> escalate immediately regardless of AI availability.
- AI call via fusion.api.service.call_openai() (consumer='fusion_repairs',
feature='client_self_triage') with try/fallback per project rule -
no hard fusion_api dep, no install error if it's missing.
- Strict response validation: JSON schema check, max 3 steps, max 200
chars per field, forbidden-phrase regex (diagnose, you have, medical
condition, stop using, consult doctor, price patterns) - on any
failure falls back to deterministic rules.
- 24h in-memory cache keyed by (category, symptom_hash) so repeat calls
during AI cost-cap incidents come from cache.
- System prompt + JSON schema published as ir.config_parameter so office
can refine without code changes (default prompt + schema in spec
Appendix A).
- New fusion.repair.self.check.rule model + 17 seeded rules across all
7 product categories (data/self_check_data.xml) - these are the
deterministic fallback AND the canonical seed if AI is disabled.
- New /repair/self_check jsonrpc route (auth=public) gated by the
per-IP rate-limit; defensive input bounds (max 5 symptoms, 500 chars
each) defend against prompt-injection bloat.
CL15 weekend safety escalation + on-call paging
- New fusion.repair.on.call.service AbstractModel with:
* find_next_on_call(exclude=...) -> lowest x_fc_on_call_priority
* page_on_call(repair) -> sends mail to next available + writes
x_fc_on_call_token / x_fc_on_call_paged_user_id / paged_at on the
repair, posts chatter
* acknowledge(repair, user) -> records ack, posts chatter
* cron_escalate_unacknowledged() -> every 5 min, re-pages the next
priority for repairs paged >15 min ago without ack
- Auto-fires from intake service whenever x_fc_urgency='safety' is
submitted. _is_business_hours() defaults to "page" when no calendar
is set or after working hours.
- New email_template_on_call_page with 4px red accent + acknowledge
CTA button linking to /repair/on-call/ack/<token>.
- /repair/on-call/ack/<token> http route (auth=user, must be the paged
manager OR any internal user) records the ack and renders confirmation.
- 5-minute cron 'Fusion Repairs: Escalate unacknowledged on-call pages'
with configurable window via fusion_repairs.on_call_escalate_minutes
(default 15).
- New repair.order fields x_fc_on_call_token, x_fc_on_call_paged_user_id,
x_fc_on_call_paged_at, x_fc_on_call_acknowledged_user_ids,
x_fc_on_call_acknowledged_at - all copy=False so duplicates start fresh.
CL17 QR sticker generator
- New fusion.repair.qr.sticker.wizard TransientModel takes a Many2many
of stock.lot records (optionally filtered by product).
- QWeb PDF report fusion_repairs.report_qr_stickers prints a 4-up
sticker sheet on letter paper: 80mm x 50mm per sticker with the
QR code (38mm), product name, serial number, and the canonical
portal URL (from web.base.url + fusion_repairs.client_portal_url).
- QR encodes /repair?sn=<serial> which the public client portal
already pre-fills via the ?sn= query param.
- Uses the qrcode library if available; renders 'QR lib missing'
placeholder otherwise so the PDF still prints.
- New menu Configuration > Generate QR Stickers + standalone wizard.
Verified end-to-end on local westin-v19:
CL6 stairlift+smoke -> escalate=True source=escalated reason=safety
CL6 bed (no AI) -> fallback returned escalate=True (safe default)
CL15 admin paged for RO-202605-10 with 27-char token
CL17 sticker URL: /repair?sn=001124032521528404
QR data URI: data:image/png;base64,iVBORw... (PNG OK)
Bumped to 19.0.1.2.0 (minor bump - new public-facing capabilities).
Co-authored-by: Cursor <cursoragent@cursor.com>
H1 Float -> Monetary for outstanding_balance
Added currency_id companion field on the wizard so widget="monetary"
renders properly. Currency defaults to env.company.currency_id.
H2 Maps URL address duplication
fusion_tasks address_street often contains the full Google-Places-
formatted address. Concatenating address_street + address_city + zip
was producing "15 Fisherman Dr, Brampton, ON L7A 1B7, Canada, Brampton,
L7A 1B7". Now uses the existing address_display field (fusion_tasks
computes it correctly for both Google Places and manual entries), with
a partner-based fallback that includes street, street2, city,
state_id.name, zip, country_id.name.
H3 Banner copy hardcoded "14 days"
Added duplicate_window_days compute field; banner now reads
"in last <N> days" from the ir.config_parameter.
H4 Outstanding-balance multi-company + child_of direction
- Dropped .sudo() (CS users already have access to their own company's
invoices via standard groups + the Repairs Office rule)
- Replaced child_of (which only walks descendants) with
commercial_partner_id (the canonical Odoo "billed-to root" - covers
child contacts AND walks up from a child if the caller IS a child)
- Added ('company_id', 'in', env.companies.ids) filter to both the
invoice search AND the duplicate-repair search so a CS rep in
Westin Healthcare doesn't see NEXA Systems balances
H5 duplicate_count capped at 5 (false reassurance)
Now uses search_count for the true total + search(limit=5) for the
display list. Earlier verification showed count=5 was actually
capped; running again shows 15 for the same partner.
M1 Function-level imports
Moved urllib.parse.quote_plus and odoo.exceptions.UserError to module
top in technician_task.py.
M2 Many2many 'in' with scalar
Changed ('x_fc_repair_skills', 'in', category.id) to
('x_fc_repair_skills', 'in', [category.id]) - safer against future
ORM tightening.
M4 C6 - added x_fc_is_quote_only field + filter + form indicator
Boolean tracked field on repair.order (was previously discoverable
only via chatter text). Indexed. Visible on the form's intake metadata
row and filterable on the dashboard search view as "Quote Only".
M5 Account-move read perf
Replaced Move.search() + Python sum with _read_group(
aggregates=['amount_residual:sum', '__count']) - pushes the SUM to
Postgres; O(1) record load vs O(N).
M6 Hide Maps button when no address
Added invisible="not address_display and not partner_id" on the
Open in Maps button so it doesn't appear on in-store tasks.
Plus the dispatch-task cutoff is now a datetime (was a date) so the
create_date >= cutoff comparison is type-correct.
Verified end-to-end on local westin-v19 after fixes:
C1 count: 15 (was capped at 5) window_days: 14
C5 balance: 0.0 currency: CAD warning: False (correct)
C6 x_fc_is_quote_only: True tech_tasks: 0 (urgent intake, NOT dispatched)
T1 URL: https://www.google.com/maps?q=15+Fisherman+Dr%2C+Brampton%2C+ON+L7A+1B7%2C+Canada%2C+Unit+7
(no duplicated city/zip)
Bumped to 19.0.1.1.1.
Co-authored-by: Cursor <cursoragent@cursor.com>
C1 duplicate-call detection
- Wizard computes duplicate_count + duplicate_repair_ids when partner is
picked (open repairs from the configurable window, default 14 days).
- Yellow banner with "Open Existing Repair" button to jump to the most
recent duplicate so CS can add a note instead of creating a new repair.
C5 outstanding-balance warning
- Wizard sums posted unpaid account.move.amount_residual across all
invoices of the partner.
- Red banner shown when balance >= fusion_repairs.outstanding_balance_threshold
(default $100) with a "View Invoices" button.
C6 quote-only mode
- New quote_only boolean on the wizard; passed through the shared intake
service. Skips dispatch-task creation for urgent/safety AND for catalogue
auto_schedule. Chatter note "Created in Quote Only mode" posted on the
resulting repair.order.
D2 skills filter on dispatch picker
- _pick_dispatch_technician(repair) prefers users whose x_fc_repair_skills
Many2many contains the repair's product category. Three-tier preference:
1) intake user if field staff AND has the skill
2) any active field-staff user with the skill
3) any active field-staff user (no skill filter) - last-resort
- Logs a warning + skips task creation if no field-staff user exists at all.
T1 Open in Maps on technician task
- action_open_in_maps() returns ir.actions.act_url to
https://www.google.com/maps?q=<URL-encoded address>. Deep-links into
Apple Maps / Google Maps native apps on iOS / Android, browser otherwise.
- Header button added on the fusion.technician.task form (after the
existing buttons) plus a "View Repair" button when x_fc_repair_order_id
is set.
Verified end-to-end on local westin-v19:
Existing repair: RO-202605-06
C1 duplicate_count = 5 (>=1 expected) - last duplicate: RO-202605-06
C5 balance check ran without error (target partner had $0)
C6 quote-only repair: RO-202605-07 tech_tasks = 0 (expected 0)
D2 picked the only stairlift-skilled field-staff user
T1 Maps URL: https://www.google.com/maps?q=15+Fisherman+Dr%2C+Brampton%2C+ON+L7A+1B7%2C+Canad...
Bumped to 19.0.1.1.0.
Co-authored-by: Cursor <cursoragent@cursor.com>
Microsoft Graph's delta API returns @removed={reason:'changed'} when an
event drifts outside the original delta-query window — the event still
exists upstream. The old code treated any truthy @removed the same as a
real delete and archived the local calendar.event. Combined with
_find_existing_event filtering by active=True, every subsequent sync
recreated a duplicate (then archived it on the next pass), accumulating
5x duplicates and emptying the user's calendar.
- _process_microsoft_event: only archive on isCancelled or
@removed.reason='deleted'; skip on @removed.reason='changed'
- _process_microsoft_event link path: reactivate when MS Graph confirms
a previously-archived event still exists
- _process_microsoft_event iCalUId path: same reactivation
- _find_existing_event: include archived records so wrongly-archived
duplicates are reused instead of piling up
- callers reactivate the matched archived record
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
When admin (gsingh, uid=2) opened a repair on the dashboard:
"Sorry, Gurpreet Singh (id=2) doesn't have 'read' access to:
- Repair Order, RO-202605-04 (repair.order: 34)
Blame the following rules:
- Repair Order: Technician sees own repairs"
Root cause: per-group record rules in Odoo are OR'd within the same
model. Admin had been added directly to fusion_tasks.group_field_technician
in this database (verified via res_groups_users_rel - direct=1), so the
technician's restrictive rule ('only repairs you are assigned to') kicked
in. Until now there was no per-group rule for the Repairs Office groups
to OR against, so the restrictive rule won by default.
Fix - added two pairs of permissive rules:
rule_repair_order_repairs_user_full - User can read/write/create
rule_repair_order_repairs_manager_unlink - Manager also can delete
rule_technician_task_repairs_office - User can read/write/create tasks
rule_technician_task_repairs_manager_unlink - Manager also can delete tasks
Both have domain_force=[(1,'=',1)] so they grant unrestricted access for
the Repairs groups. OR'd with the field_technician rule, admin and other
office users now see everything. Field technicians who do NOT have any
Repairs group still see only their assigned repairs (rule unchanged).
Also added the matching ir.model.access.csv entries - record rules don't
fire if the user has no model-level ACL. This is the second fix
('office users can schedule') from the same complaint - Repairs User now
has read/write/create on fusion.technician.task; Repairs Manager also
gets unlink.
Verified end-to-end on westin-v19:
Admin can see 17 repairs (was 0 before fix)
Admin can read RO-202605-04 -> 'Gurpreet Singh' (the exact failing record)
Admin can create fusion.technician.task -> permission check passes
(model's own time-overlap business validation correctly rejects an
overlap, but that is a value error not a permission error)
Bumped to 19.0.1.0.7.
Co-authored-by: Cursor <cursoragent@cursor.com>
The dashboard root used min-height: calc(100vh - 46px) which expanded
to the viewport but bypassed the parent .o_action_manager flex sizing,
so the inner overflow-y: auto had nothing to scroll - vertical content
was clipped or stuck.
Replaced with height: 100% + overflow-y: auto + overflow-x: hidden so
the component fills its action container and scrolls naturally. Bumped
to 19.0.1.0.6 to bust the asset bundle hash.
Co-authored-by: Cursor <cursoragent@cursor.com>
A real landing dashboard for the Fusion Repairs app so users see at a
glance what is open, what is urgent, and where to click. Built as an
OWL client action, theme-aware (light AND dark) at SCSS compile time,
zero hardcoded user-facing colours.
What's on it
- Hero banner with gradient accent
- 4 quick-action tiles (New Service Call, Service Calls, Maintenance
Contracts, Repair Warranties)
- 6 KPI stat tiles (Open / Urgent+Safety / Awaiting Dispatch /
Needs Re-Quote / New This Month / Maintenance Due 30d) - each is
clickable and lands in the right filtered list
- Self-service portal cards with copy-to-clipboard for the public
client portal URL and the sales rep portal URL (so office can
share them on voicemail / printed materials / training)
- Recent Service Calls list (last 5) - click jumps to repair form
- Upcoming Maintenance list (next 5 due) - red pill when <=7 days out
- Configuration tiles (Equipment Categories / Intake Templates /
Service Catalogue)
- Refresh button
Architecture
- fusion.repair.dashboard AbstractModel exposes get_dashboard_data():
returns stats + urgency_breakdown + source_breakdown + recent[5] +
upcoming[5] + portals (URLs resolved via web.base.url +
fusion_repairs.client_portal_url)
- FusionRepairsDashboard OWL component (registry actions
'fusion_repairs.dashboard') uses standalone rpc() per project rule
#3, useService('action') for navigation, useService('notification')
for copy feedback. static props = ['*'] to accept the client-action
props envelope.
- _fr_tokens.scss registered FIRST in web.assets_backend so its
variables are in scope when dashboard.scss compiles. NO @import (per
project rule). Branches on $o-webclient-color-scheme at compile time
so the dark bundle (web.assets_web_dark) gets dark hex values
automatically - per project CLAUDE.md rule on dark mode.
- All visible colours come from CSS-variable-wrapped SCSS tokens
(--fr-page-bg, --fr-card-bg, --fr-border, --fr-accent, ...) which
fall back to the SCSS hex value. Three-layer contrast: page (grayest)
-> card (mid) -> elevated (brightest).
- New ir.actions.client action_fusion_repairs_home_dashboard with
tag='fusion_repairs.dashboard'.
- Top-level menu now lands on this dashboard. 'Dashboard' added as
the first sub-menu; 'Service Calls' (the kanban) is still right
below it.
Verified on local westin-v19:
STATS: open=15, urgent=4, new_this_month=13, awaiting_dispatch=9,
requires_requote=1, maintenance_due_30d=1, active_total=2
PORTALS: client=http://192.168.139.165:8069/repair
sales_rep=http://192.168.139.165:8069/my/repair/new
RECENT count: 5
UPCOMING count: 2
SOURCE breakdown: backend_wizard 9, client_portal 3, manual 2, sales_rep_portal 1
Web /web/login: 200, no SCSS compile errors in logs.
Bumped to 19.0.1.0.5 so the asset bundle hash refreshes.
Co-authored-by: Cursor <cursoragent@cursor.com>
Replaced the picking-type default reference (BR-WA/RO/00010) with a
date-based monthly-resetting sequence: RO-202605-01, RO-202605-02, ...
where YYYY is the year and MM is the zero-padded month. The counter
resets to 01 every time the month rolls over.
Implementation:
- New ir.sequence 'fusion.repair.order.monthly' with prefix
'RO-%(year)s%(month)s-', padding=2, use_date_range=True (Odoo creates
one ir.sequence.date_range per month, each with its own number_next)
- repair.order.create() override pre-fills vals['name'] with the new
sequence BEFORE super(), so Odoo's native picking-type sequence
assignment (which only fires when name is empty / 'New') is bypassed
Verified on local westin-v19: three back-to-back creates produced
RO-202605-01 / -02 / -03. Existing records (pre-upgrade) keep their
old BR-WA/RO/##### references - this only affects repairs created
from this version onward.
Bumped to 19.0.1.0.4.
Co-authored-by: Cursor <cursoragent@cursor.com>
Reports of literal '<b>Client Self-Service</b>' showing in the chatter
instead of bold formatting. Cause: message_post(body=str) HTML-escapes
the string. The Odoo idiom for HTML chatter bodies is markupsafe.Markup,
with the % operator auto-escaping substitution values for XSS safety.
Fixed every message_post call:
models/intake_service.py
- 'Service call submitted via <b>...</b>' (the reported one)
- 'This repair MAY be covered by our active warranty <b>...</b>'
models/maintenance_contract.py
- 'Sent N-day maintenance reminder to <email>'
- 'Maintenance visit <b>...</b> booked from reminder link'
models/technician_task.py
- 'Rolled forward after maintenance task <b>...</b> completed'
wizard/repair_visit_report_wizard.py
- 'Spawned follow-up repair <b>...</b> for "found another issue"'
Pattern used: Markup(_('... <b>%(x)s</b> ...')) % {'x': escaped_value}.
Verified on local westin-v19 (BR-WA/RO/00026): DB row now reads
'<p>Service call submitted via <b>Client Self-Service</b> by Gurpreet
Singh. Session reference: RIS000015.</p>' which renders correctly in
the chatter UI.
Bumped to 19.0.1.0.3.
Co-authored-by: Cursor <cursoragent@cursor.com>
Two complaints from the first hands-on test:
1) Submit button raised "Access Error (Document type: Activity,
Operation: create)" - the wizard called the intake service WITHOUT
sudo so the mail.activity records the service schedules tripped on
the activity ACL (admin's group chain does not auto-grant activity
create on repair.order without sudo). Both portal controllers
already sudo'd; the wizard now does too. x_fc_intake_user_id
preserves audit identity regardless.
Verified end-to-end as gsingh@westinhealthcare.com (admin):
Created: BR-WA/RO/00025
Activities: 2
Source: backend_wizard
Intake user: gsingh@westinhealthcare.com
2) "Real dashboard with dedicated pages would have been nice" - the
main menu opened the wizard directly as a modal. Restructured so
the menu lands on a proper kanban dashboard of service calls,
matching the standard Odoo app pattern:
Fusion Repairs (app icon)
- Service Calls <- dashboard kanban (default landing)
- New Service Call <- wizard (still a modal, accessed from menu OR kanban's New button)
- All Repair Orders <- native Odoo repair list (full backend)
- Maintenance Contracts
- Configuration
- Equipment Categories / Intake Templates / Service Catalogue / Repair Warranties
New view_fusion_repair_dashboard_kanban shows urgency badges (red /
amber / grey), category, scheduled date, intake source pill, and
a 3rd-party warning. Default group_by=state.
New view_fusion_repair_dashboard_search adds quick filters: Today,
This Week, Safety/Urgent, Third-Party, Open, plus per-source filters
and Group By (Status / Urgency / Category / Intake Source).
Wizard remains target='new' (modal) so submitting drops the user
back to the kanban they came from with the new repair visible.
Bumped version to 19.0.1.0.2 to bust the asset bundle hash.
Co-authored-by: Cursor <cursoragent@cursor.com>
Two related issues that hid the Fusion Repairs app from the Apps menu
for admin users:
1. Custom security groups don't auto-include admin
The Repairs User / Dispatcher / Manager groups are new custom groups.
Having base.group_user or base.group_system on its own does NOT grant
membership in custom child groups - implied chains only flow one way
(child -> parent). Admin therefore had no Repairs groups, so the
top-level "Fusion Repairs" menu (gated on group_fusion_repairs_user)
was hidden from them.
Fix: extend base.group_system with implied_ids that include
group_fusion_repairs_manager. Manager already implies Dispatcher
implies User, so admin (= base.group_system) now automatically gets
the whole chain on install / upgrade with no manual user editing.
Verified via odoo-shell:
admin.has_group('fusion_repairs.group_fusion_repairs_user') == True
admin.has_group('fusion_repairs.group_fusion_repairs_dispatcher') == True
admin.has_group('fusion_repairs.group_fusion_repairs_manager') == True
menu_fusion_repairs_root._filter_visible_menus() == ir.ui.menu(2735,)
2. Missing static/description/icon.png
The manifest referenced fusion_repairs,static/description/icon.png
via web_icon on the top-level menu but the file did not exist. Odoo
handles missing icons gracefully but the apps list ends up rendering
without a tile graphic. Copied fusion_tasks/static/description/icon.png
as a placeholder; replace with a custom asset whenever desired.
Verified: /fusion_repairs/static/description/icon.png returns
HTTP 200 with 43989 bytes after restart.
Bumped manifest version to 19.0.1.0.1 to bust the asset bundle hash so
clients pick up the new icon without a manual cache clear.
Co-authored-by: Cursor <cursoragent@cursor.com>
The cross-instance sync silently drops tasks when x_fc_tech_sync_id is
missing on the technician, and silently collapses duplicates via dict
comprehension. Both make sync break in ways that are invisible until
someone notices a missing task on the other instance.
- _get_remote_tech_map / _get_local_syncid_to_uid: warn on duplicates
- _push_tasks_to_remote: info-log when a task is skipped because the
tech has no sync_id or no remote counterpart
- res.users onchange: warn in the form when entering a sync_id that
is already used by another active field staff
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Critical
- C1: _sql_constraints -> models.Constraint (Odoo 19 deprecation rule violation)
- C2: variance threshold no longer uses abs() - under-cost is good news,
must not block invoicing. Now only OVER-cost triggers requires_requote.
- C3: roll_next_due_date() was dead code - now wired from
fusion.technician.task.write() when a maintenance task transitions to
'completed', so the whole maintenance lifecycle actually advances.
- C4: warranty.is_active was store=True but time-dependent (became stale).
Dropped store=True; find_active_for() now filters by expiry_date directly.
High
- H1: added x_fc_maintenance_contract_id back-link on repair.order and
populated it from create_repair_from_booking().
- H2: find_active_for() returns empty when neither lot nor product is
supplied - prevents cross-product false warranty matches.
- H3: visit-report wizard now creates stock.move records of repair_line_type
'add' for each part line, so Odoo's native action_create_sale_order()
chain has lines to invoice and stock gets consumed properly.
- H4: office intake email template now carries a fallback email_to header
computed from res.company.x_fc_office_notification_ids (or company email),
so it does not silently send with no recipient.
- H5: maintenance reminder cron nextcall now always rolls to tomorrow
at 07:00 local time, so installing/upgrading after 07:00 does not
immediately fire all the day's reminders.
- H6: public portal no longer hardcodes UID 1 as the intake user fallback
(which in Odoo 19 is OdooBot). Prefers base.user_admin, else the
lowest-id non-share user, else SUPERUSER_ID.
- H7: public portal validates client_email via tools.email_normalize
before partner creation; malformed addresses redirect with error=email.
- H8: find_best_match() returns empty when no symptom keywords match
(no silent first-catalog guess) and uses word-boundary regex to avoid
matching 'battery' inside 'no battery problem'.
Medium
- M1: _inherit moved next to _name on maintenance_contract (cosmetic but
brittle if Odoo refactors model class detection)
- M2: relativedelta(months=N) instead of timedelta(days=N*30) for warranty
and maintenance intervals (correct month boundaries)
- M3: unique constraint on fusion.repair.maintenance.contract.booking_token
- M6: dispatch task fallback now searches for an actual x_fc_is_field_staff
user; gracefully skips and logs if no field staff exists (instead of
silently failing the constraint check)
- M7: maintenance contract list view date decoration uses context_today()
(date) instead of strftime(string) - the str comparison would TypeError
- M9: Visit Report button hidden on draft repairs and when no technician
task is linked yet
Low
- L2: portal-created partners get default lang + company_id so mail
templates render in the right language
- L3: dropped unused exception variable in sales rep portal controller
- L4: visit-report wizard 'found another issue' now redirects to the
spawned stub repair so the tech can fill it in immediately
- L5: dropped unrecognized data-string from <app> in settings view
Public portal also: rate-limit check moved BEFORE the counter increment so
blocked attempts do not keep inflating the bucket.
All fixes verified end-to-end on local westin-v19:
- variance one-sided: 0.5h labour vs $500 est -> requires_requote=False;
2h x $250 + $200 parts vs $100 est -> requires_requote=True
- maintenance roll-forward: created MC/00006 due 2026-05-31, completed
linked maintenance task -> contract rolled to 2026-11-21 with
last_reminder_band reset
- warranty find_active_for(partner only) -> empty recordset
- service catalog find_best_match with unrelated text -> empty recordset
- pg_constraint shows fusion_repair_maintenance_contract_booking_token_unique
- /repair landing still 200 after restart
Co-authored-by: Cursor <cursoragent@cursor.com>
On the original purchase sale.order:
- Repairs button (fa-wrench) lists all repair.order records where
x_fc_original_sale_order_id = this SO
- Maintenance button (fa-calendar-check-o) lists all
fusion.repair.maintenance.contract records spawned from this SO
- Both auto-hide when count is zero
- Both gated by fusion_repairs.group_fusion_repairs_user
Follows the count + action_view_* + oe_stat_button / statinfo pattern
from fusion_claims/views/sale_order_views.xml line ~1176.
Co-authored-by: Cursor <cursoragent@cursor.com>
Maintenance contracts
- New fusion.repair.maintenance.contract model: one per partner +
product + lot. Fields: interval_months, last_service_date,
next_due_date, state, booking_token (secrets.token_urlsafe),
last_reminder_band (30 / 7 / 1), booking_repair_id
- roll_next_due_date() advances the cycle by interval_months and resets
the band / booked-repair so the next cycle starts fresh
- sale.order._spawn_maintenance_contracts() creates contracts for
delivered SOs whose product has x_fc_maintenance_interval_months > 0
(called from Phase 3 hooks; ready for cron / on-state change wiring)
Reminder cron
- Daily ir.cron at 07:00 -> cron_send_due_reminders()
- Sends email at 30 / 7 / 1 day bands before next_due_date; tracks
last_reminder_band so we never re-send the same band in one cycle
- Master toggle via ir.config_parameter fusion_repairs.enable_email_notifications
Public client booking portal
- /repairs/maintenance/book/<token> GET landing page with a date input
- /repairs/maintenance/book/<token>/confirm POST creates a repair.order
via contract.create_repair_from_booking() (source='client_portal')
- Idempotent: existing booking shows "already booked" instead of
spawning a duplicate
- Invalid / expired tokens render a friendly "link not valid" page
Mail template
- email_template_maintenance_due_reminder with 4px green accent bar,
600px max-width, dark/light safe; renders the tokenized booking CTA
button directly to /repairs/maintenance/book/<token>
Backend
- Maintenance Contracts list / form with statusbar + chatter
- Menu under Operations -> Maintenance Contracts
- Sequence MC/##### for contract reference
- Access rules: User read, Dispatcher write, Manager full
Verified end-to-end on local westin-v19:
- Contract MC/00003 created due in 7 days
- cron_send_due_reminders() fires the 7-day band; second invocation
skips (idempotent)
- create_repair_from_booking() spawns BR-WA/RO/00014 with
x_fc_intake_source='client_portal' and links it back to the contract
- HTTP GET /repairs/maintenance/book/<token> -> 200 with the date input
and contract reference visible in the page
Co-authored-by: Cursor <cursoragent@cursor.com>
Service catalogue
- New fusion.repair.service.catalog model: named service entries per
equipment category with symptom keywords, estimated hours / cost,
default parts, auto_schedule flag, optional pricelist override
- find_best_match() scores candidates by symptom-keyword overlap against
intake text hints (issue summary + category + notes)
- Intake service wires it in: on submit, the matcher sets
x_fc_service_catalog_id + x_fc_estimated_duration + x_fc_estimated_cost
and (when auto_schedule=True) creates a draft dispatch task
- Double-task guard: if catalogue match already created a task, the
urgency-based dispatch skips so we never duplicate
Visit report wizard
- fusion.repair.visit.report.wizard with labour hours + parts lines +
technician notes + 'found another issue' branch
- Computes actual cost = (labour x service_product.list_price) + parts
- Compares against estimate -> sets requires_requote when variance
exceeds configured threshold (% or $); shows warning banner inline
- On confirm: writes actuals back to repair, posts notes to chatter,
optionally spawns a follow-up repair (T5 'found another issue')
Repair warranty
- New fusion.repair.warranty.coverage model (start/expiry, partner,
product, lot, active flag)
- find_active_for(partner, product, lot) returns the most-recent active
coverage
- Intake service auto-checks: when a new repair lands on an equipment
that has active warranty coverage, posts a chatter banner so the
office knows the work may be free under our 30/90-day re-do policy
(manager review still required; never auto-zeros pricing)
Repair form
- Header: Visit Report + Collect Payment buttons (gated by group)
- action_collect_payment looks up the linked posted unpaid invoice on
the repair SO and opens the Poynt wizard (action_open_poynt_payment_wizard)
AI intake summary
- _generate_ai_summary calls self.env['fusion.api.service'].call_openai
with consumer='fusion_repairs', feature='intake_triage'
- Strict system prompt: no medical advice, no diagnoses, no recommending
stop equipment use; ~80 words; plain English
- Try/fallback per fusion-api-integration.mdc: if fusion_api not
installed or call fails -> silently skip; intake never blocked
Verified end-to-end on local westin-v19:
- Stairlift motor intake -> catalogue match -> estimated $500/2h -> auto
dispatch task (count=1, not duplicated)
- Visit report: 2.5h x $250 + $100 parts = $725 actual vs $500 estimated
= 45% variance -> requires_requote=True
- Warranty: 30-day coverage on the completed repair; second repair on
same partner triggers warranty banner in chatter
Co-authored-by: Cursor <cursoragent@cursor.com>
Both portals share the existing fusion.repair.intake.service so behaviour
stays identical across all three intake surfaces (backend wizard,
sales rep portal, public client portal).
Sales rep portal
- Hard depends on fusion_authorizer_portal (reuses is_sales_rep_portal
flag + group_sales_rep_portal scaffolding)
- /my/repair/new - mobile-friendly intake form with phone-first
partner search (jsonrpc lookup), category select, third-party flag,
urgency, photo capture
- /my/repairs - list of repairs the rep submitted (paginated)
- /my/repair/<id> - read-only detail with status, equipment, scheduled
visit
- Interaction-class JS (Odoo 19 public.interactions), safe DOM construction
- Mobile SCSS with 44px tap targets, sticky CTA on small screens
- Record rule scopes portal users to repairs where
x_fc_intake_user_id = user.id
Public client portal
- auth='public' - voicemail-ready /repair URL
- /repair - landing page with 911 disclaimer and Start CTA
- /repair/new - single-page form: contact, equipment, issue, urgency,
optional photos. QR pre-fill via ?sn=<serial>
- /repair/submit - CSRF + honeypot + per-IP rate limit (configurable);
finds or creates partner; calls intake service with sudo
- /repair/thanks - confirmation with reference number
- /repair/lookup_phone (jsonrpc) - safe partner match returning ONLY
masked name (first + last initial) + city (no other PII leakage)
Security fix: technician record rule on repair.order now uses STORED
fields (technician_id + additional_technician_ids) instead of the
non-stored all_technician_ids compute, which was failing SQL generation.
Verified end-to-end on local westin-v19:
- Sales rep create via intake service with the rep user context creates
the repair with x_fc_intake_source='sales_rep_portal' and proper
activities
- /repair/submit posts urlencoded data -> creates partner + repair
('BR-WA/RO/00010', source='client_portal', urgency='urgent') ->
redirects to /repair/thanks with the reference
Co-authored-by: Cursor <cursoragent@cursor.com>
Comprehensive 4-phase design for fusion_repairs Odoo 19 module covering
three intake surfaces (backend wizard, sales rep portal, public client
portal), AI self-check with strict medical safety guardrails, weekend
on-call paging, repairs pricelist automation, Poynt payment collection,
and maintenance lifecycle with client self-booking. 53 features across
phases 1-4; reuses existing fusion_tasks technician model and
fusion_authorizer_portal sales rep scaffolding.
Includes Appendices A-D with seed AI system prompt + JSON schema,
15 upsell rules, voicemail scripts, and 30 deterministic self-check
rules across 7 medical equipment categories.
Co-authored-by: Cursor <cursoragent@cursor.com>
Previous attempt (e5928b96) used CSS to force the binary widget's
"Upload your file" button visible in display mode. Problem: it
rendered a non-clickable stub in every row, then DUPLICATED when
the operator clicked into edit mode (two upload links stacked).
Drop the SCSS hack entirely. Replace with a custom form-view
controller that auto-edits the first incomplete row on mount.
When the wizard opens, the JS:
1. Scopes itself via the form's o_fp_cert_issue_wizard_form class
(no-ops on every other form view in the system).
2. Finds rows where the is_ready toggle is False.
3. Clicks the fischer_file cell of the first such row.
4. The row enters edit mode → Odoo's native binary widget renders
its upload button → operator drops the file → onchange fires
→ readings parse.
Wired via js_class="fp_cert_issue_wizard_form" on the form root.
Banner copy updated to "Click a row, then click Upload your file in
the Fischerscope column" so even if the auto-edit fails for some
DOM reason, the operator knows the click path.
Module: fusion_plating_jobs 19.0.10.16.1 → 19.0.10.16.2.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Reported 2026-05-20: the Fischerscope file column shows "↑ Upload
your file" only when the operator clicks the cell. Until then, the
cell looks empty and operators don't know they can upload there.
Root cause: Odoo's default `widget="binary"` only renders the
upload button in EDIT mode. In editable lists, non-selected rows
stay in display mode, which hides the button. Stock theme CSS
hides .o_select_file_button on inactive rows.
Fix: scoped SCSS that overrides the default theme rule for the
Issue Certs wizard ONLY. `.o_select_file_button` becomes
`display: inline-flex !important` so it shows on every row from
the moment the wizard opens. Added a fa-upload icon glyph + dotted
underline so the button reads as clickable-action, not text.
Scoped to `.o_field_one2many[name="line_ids"]` inside the form view
so binary fields elsewhere in the system are unaffected. Registered
in both web.assets_backend and web.assets_web_dark per CLAUDE.md
two-bundle rule.
Module: fusion_plating_jobs 19.0.10.16.0 → 19.0.10.16.1.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Reported 2026-05-20: clicking "Issue Cert" on a job opened the
wizard with a banner saying "Fischerscope file or readings needed
— fill it in below before confirming", but the list view only
showed status toggles (Needs Thickness / Is Ready). No upload
affordance was visible. Operators had to know they could click a
list row to expand into a hidden detail form where the upload
field lived.
The wizard model already had the file field, the .docx parser
(_fp_parse_fischerscope_docx), and the @onchange that prefills
readings — only the view was hiding it.
Fix: promote the file upload into the list as its own editable
binary column, alongside the existing Needs Thickness toggle.
Operator now sees:
Reference │ Type │ Customer │ Needs Thickness │
Fischerscope File (PDF or .docx) │ Parsed │ Ready
Drop the file → onchange fires → readings + parsed summary
populate in-row. Click "Confirm & Issue" to commit.
The per-line expanded form is preserved (still accessible via
row click) as a "details" panel for editing individual readings
after upload — but the primary upload action is now in the list
row where the operator's eyes are.
Module: fusion_plating_jobs 19.0.10.15.0 → 19.0.10.16.0.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Reported 2026-05-20: the receiving state machine had four states
(draft → counted → staged → closed) where the middle pair was pure
ceremony. Real-usage data on entech:
state distribution: 14 draft, 4 closed (zero `staged` records)
median dwell counted → staged: 11 seconds
median dwell staged → closed: 4 minutes
`staged` captured no fields, fired no gates, mapped to the same SO
`x_fc_receiving_status='partial'` as `counted`. Pure click-through.
Cleanup:
- State Selection retains `staged` as `Staged (legacy)` so historical
records remain readable; new transitions never write it.
- statusbar_visible drops it from the chevron header.
- action_mark_staged becomes a thin shim that advances counted →
closed directly (any old button binding still works).
- action_close now accepts `counted` as a valid source state (was
previously only `staged` / legacy `accepted` / `resolved`).
- View: "Stage for Racking" button removed. "Close" button renamed
to "Close — Racking Confirmed" so the racking-crew confirmation
meaning stays obvious.
- _update_so_receiving_status mapping unchanged for legacy `staged`
(still maps to partial) — only the comment block updated to
describe the new canonical flow.
Migration 19.0.3.20.0 advances any `staged` records to `closed`
and syncs the linked SO's x_fc_receiving_status to `received` so
downstream gates (job step start, mark_done qty check, cert
creation) don't see a stale "partial" status.
Module: fusion_plating_receiving 19.0.3.19.0 → 19.0.3.20.0.
Tests: TestQtyReceivedPropagation updated — 5 tests dropped the
action_mark_staged() call, walk draft → counted → closed directly.
All 11 tests green (carrier 6 + propagation 5).
Verified on entech: existing 14 draft + 4 closed records untouched.
Direct draft → counted → closed transition works end-to-end on
RCV-30041 (was the test target).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Reported 2026-05-20: the Template dropdown in the Part > Process
Composer's 'Add Variant from Template' row truncated long recipe
names to 4 characters ("Cher" instead of "Chemical Conversion …").
The hard-coded max-width: 280px was set before the curated template
catalog grew names like "Chemical Conversion — Iridite Type II Cl 3"
and "ENP-STEEL-BASIC — Standard Heavy Phos".
Fix: replace the rigid max-width with a flex sizing that gives the
dropdown room to grow:
- min-width: 360px (full common recipe name fits)
- flex: 1 1 360px (grows to fill available space)
- max-width: 560px (cap so it doesn't push the buttons off-screen)
Same flex pattern applied to the Variant label input (slightly
narrower min/max).
Also: pulled the entech-side version of fp_part_process_composer.xml
back into the local repo — local was stale (one 'Add Variant' button;
entech had the dual 'Add — Tree' / 'Add — Simple' buttons that
landed in an out-of-band edit).
Module: fusion_plating_configurator 19.0.21.5.0 → 19.0.21.5.1.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Reported 2026-05-20: on a 40+ step recipe (e.g. ENP-STEEL-BASIC),
scrolling down into the Selected steps pane scrolled the Step
Library off the top of the screen. Authors had to scroll back up
to grab a step, then scroll down to drop it.
Fix: position: sticky on .o_fp_library_panel, pinned to top: 1rem
(matches the editor's padding) inside the .o_fp_simple_editor
overflow container. align-items: start on the grid so the library
column doesn't stretch to match the recipe column's height
(prerequisite for sticky to behave).
The library itself can have 30+ entries (curated step kinds +
shop-defined library templates). max-height: calc(100vh - 8rem)
+ overflow-y: auto keeps it from blowing past the viewport — it
grows its own internal scrollbar instead.
Mobile (≤900px) reverts to static positioning so the stacked
layout stays sensible.
Module: fusion_plating 19.0.20.6.1 → 19.0.20.6.2.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Regression of an earlier fix. Operators reported the editor jumping
to the top of the page on every step save / insert / remove / promote.
Root cause: .o_fp_simple_editor is the overflow:auto scroll
container. loadAll() replaces state.steps with a fresh JSONRPC
payload — OWL tears down the t-foreach and rebuilds every row, which
snaps scrollTop back to 0. Every author action (Save Step, Add
Step, Remove, Promote, Demote, Reorder, Import Template) routes
through loadAll, so the symptom hit everywhere.
Fix: capture scrollTop before the RPC, restore in a double-rAF
after the response settles. rAF (microtask runs before paint in
OWL 2; we need the rebuilt DOM to exist). One choke point fix —
every caller benefits without per-handler changes.
Cheap: a single DOM lookup + an integer save/restore. No XML or
state-shape changes.
Module: fusion_plating 19.0.20.6.0 → 19.0.20.6.1.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Operator-reported foot-gun: Step Kind dropdown had 24 options, most
of which were visual-only (cleaning, electroclean, etch, rinse,
strike, dry, wbf_test, hardness_test, adhesion_test, salt_spray,
packaging, etc.) and didn't drive any gate or milestone. Picking the
wrong one meant nothing happened; picking Generic (left default)
meant nothing happened. Authors couldn't tell which choice mattered.
Curation: 24 → 11 active kinds. Each remaining kind has a concrete
downstream behaviour (gate, portal milestone, hardware tie-in, or
"explicitly no behaviour" for Other):
other Other (catch-all, default — no special behaviour)
receiving Received portal milestone
contract_review QA-005 form gate + button_finish lock
racking Rack-assignment dialog + button_finish lock
mask Visual mask kind (covers Masking + De-Masking)
wet_process Visual wet kind (NEW, covers cleaning, rinse,
etch, strike, dry, electroclean, wbf_test)
plate Plated portal milestone (last plate step closes)
bake Bake-window state machine + Baked milestone
inspect Intermediate inspection milestone
final_inspect Inspected (terminal) portal milestone
ship Shipped milestone (back-compat; delivery-state
driven is preferred)
Retired kinds (active=False, hidden from dropdown): cleaning,
electroclean, etch, rinse, strike, dry, wbf_test, demask, derack,
replenishment, hardness_test, adhesion_test, salt_spray, packaging,
gating. Kept in DB for audit / history but not selectable.
Mandatory enforcement:
- fp.step.kind_id on fusion.plating.process.node and fp.step.template
is now required=True with ondelete='restrict' and a default that
resolves to the 'other' kind. Existing NULL rows are backfilled by
the pre-migrate before the NOT NULL constraint hits the schema.
- Dropdown no longer offers a blank / "Generic" option. New steps
land on 'other' instead of NULL.
Admin-only catalog:
- /fp/simple_recipe/kinds/create endpoint now refuses requests from
non-managers (group_fusion_plating_manager). Returns a clear
message explaining why ("each kind drives gates / milestones /
routing — pick Other if none fits, or ask a manager to wire up a
new kind").
- "+ Add a new kind…" sentinel option in the library form is hidden
unless state.recipe.user_is_manager. Backend gate is the authority;
the UI hide is just to stop showing a button that will error.
- The Step Type dropdown in the inline step-edit panel switched from
a 24-line hard-coded XML option list to a t-foreach over
state.kindOptions (the same kinds/list endpoint payload). One
source of truth — retire / add a kind in the catalog and every
picker reflects the change.
Migration impact (entech): 5 templates + 579 nodes backfilled via
name-match heuristic. 15 kinds flipped to active=False. Distribution
of the 579 backfilled nodes:
racking 105, other 97, bake 91, wet_process 90, mask 74,
inspect 44, plate 32, final_inspect 25, receiving 10,
contract_review 9, ship 2.
Drive-by:
- Migration uses _ensure_kind() that also registers ir.model.data
for the new xmlids so the subsequent data XML load doesn't create
duplicate kind records.
- Stored related default_kind on fusion.plating.process.node /
fp.step.template is written alongside kind_id in every SQL UPDATE
so legacy `node.default_kind == 'foo'` comparisons stay accurate
(the ORM doesn't recompute stored related fields after direct
SQL writes).
Module: fusion_plating 19.0.20.5.0 → 19.0.20.6.0.
15 existing tests still green.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Three bugs reported on 2026-05-20:
1. RESURRECTION. User deletes a substep in the Simple Editor (e.g.
Soak Clean (S-3) under Cleaner), then on the next -u fusion_plating
the substep comes back. Root cause: the recipe XML lived in the
manifest's `data` list with `noupdate="1"`. Odoo's noupdate=1 only
blocks UPDATE of existing records — when a record's ir.model.data
row is missing, the loader treats it as "not yet created" and
re-creates from XML. Every upgrade resurrected every user-deleted
seed node.
Fix: pull the recipe XML files out of `data` and load them once
via post_init_hook → _seed_starter_recipes_once. Sentinel checks
ir.model.data for each recipe's root xmlid; if present, skip
loading entirely. Result: deletions are permanent across all
future upgrades. Existing entech recipes untouched.
Files affected: fp_recipe_enp_alum_basic, fp_recipe_enp_steel_basic,
fp_recipe_enp_sp, fp_recipe_general_processing, fp_recipe_anodize,
fp_recipe_chem_conversion.
2. PROMOTE / DEMOTE. Simple Editor had no way to turn a substep into
a top-level operation, or to tuck an operation under another as a
substep. Authors had to delete + re-create. New endpoints:
* /fp/simple_recipe/step/promote → flips node_type 'step' →
'operation', re-parents to the recipe (or sub-process) root,
places right after the old parent operation.
* /fp/simple_recipe/step/demote → flips 'operation' → 'step',
re-parents under the preceding operation (or a caller-supplied
target_op_id). Blocks demoting an operation that has its own
children, with a helpful message.
UI: each row in the editor now carries an up-arrow (promote, only
shown on substeps) and a down-arrow (demote, only shown on
operations). Confirmation dialog explains what's about to happen.
3. DRAG SUBSTEPS. Last commit (2142a66b) disabled drag on substep
rows. Operators couldn't reorder substeps within an operation.
Re-enabled drag on substeps. The step_reorder endpoint now groups
incoming node_ids by parent_id and renumbers within each parent
(10, 20, 30…). Cross-parent drag still no-ops on parent change —
Promote/Demote buttons are the way to move between parents.
Drive-by:
- Added `from odoo import _` to the controller (missing import the
new endpoints surfaced).
- Edit-panel field wiring audited: all fields visible in the screen
(Step name, Default instructions, Step Type, Triggers Workflow,
Parallel Start, QA Sign-off, Collect measurements, Instruction
Images, custom prompts) persist correctly through step_write or
dedicated endpoints. No broken wires.
Tests: 15 total in TestSimpleRecipeFlatten (was 10). 5 new cover
promote happy-path, promote reject (non-substep), demote happy-path,
demote block on has_children, and reorder parent-scoping.
Module: fusion_plating 19.0.20.4.0 → 19.0.20.5.0.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Follow-up to 821e768b. The previous fix flattened sub_process nodes
so all 16 operations of ENP-STEEL-BASIC became visible — but the
Tree Editor also shows the 26 `step` nodes that live under each
operation ("Ready For Blast / Blast", "Soak Clean / Electroclean /
Primary Rinse", etc.). The Simple Editor still hid those, so author
+ Tree Editor still disagreed by 26 rows.
New `_flatten_recipe_nodes(recipe)` helper walks DFS and surfaces
BOTH operations and their step children. Each operation is followed
immediately by its step children in sequence order so the editor
renders them as a contiguous block:
10. Ready For Steel Line
11. Cleaner [Steel Line]
↳ Soak Clean (S-3) [Steel Line › Cleaner]
↳ Electroclean (S-3) [Steel Line › Cleaner]
↳ Primary Rinse (S-4) [Steel Line › Cleaner]
15. Acid Dip (S-5) [Steel Line]
↳ Primary Rinse (S-6) [Steel Line › Acid Dip (S-5)]
...
Payload additions on each step:
- `node_type`: 'operation' | 'step'
- `is_substep`: True for steps (renders indented)
- `nested_under`: chained path (sub-process › operation for substeps,
sub-process for nested operations, '' for top-level operations)
UI: substep rows are indented 2.5rem, smaller font, no drag handle,
no numeric position. The "↳" indent glyph and a "[parent operation]"
chip make the parent-child relationship obvious. Substeps are not
draggable to keep the existing reorder semantics simple — Tree Editor
remains the home for structural changes.
Legacy `_flatten_recipe_operations` helper retained for back-compat
(it now delegates by filtering `node.node_type == 'operation'` from
the full walk).
ENP-STEEL-BASIC on entech: Simple Editor now shows 42 rows (was 10
before 821e768b, was 16 after 821e768b) — matches what the Tree
Editor displays exactly.
Tests: 10 total (was 7), 3 new cover the substep surfacing, path
chaining, and is_substep / node_type flags on the payload.
Module: fusion_plating 19.0.20.3.0 → 19.0.20.4.0.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Bug on ENP-STEEL-BASIC (2026-05-20): authoring used the Tree Editor
to build a recipe with a "Steel Line" sub_process holding 7 nested
operations (Cleaner, Acid Dip, Nickel Strike, E-Nickel Plate, etc.).
The Simple Editor's /fp/simple_recipe/load endpoint only walked
`recipe.child_ids`, so it returned 10 steps. The work order generator
(fp.job._generate_steps) walked the same tree depth-first and emitted
16 steps. Author and operator disagreed about what was in the recipe.
Fix: new `_flatten_recipe_operations(recipe)` helper walks the tree
depth-first, recurses into `recipe` and `sub_process`, emits each
`operation` exactly once, skips `step` children (they're sub-
instructions of operations). Mirrors the WO walker.
Step payload now carries a `nested_under` string — the chained sub-
process name(s) the operation lives inside (empty for top-level).
The Simple Editor XML renders that as a small "↳ Steel Line" badge
next to the step name so the author can see where each row came from
in the tree. Deep nesting chains with ' › ' (e.g. "Outer › Inner").
`step` children of `recipe` itself remain invisible — they were
silently skipped by the WO generator pre-19.0.18.8.0 anyway (only
operation nodes spawn fp.job.step rows). Restoring them here would
contradict that long-standing contract.
Edit/insert/reorder/remove endpoints unchanged: editing a nested
operation's name / description / tanks works (no parent change).
Drag-reorder within sub-process siblings still works. Drag across
sub-process boundaries isn't supported — opens the door for a Tree
Editor follow-up if needed, but the immediate "I can't see my
steps" complaint is resolved.
ENP-STEEL-BASIC on entech now shows all 16 operations in the Simple
Editor (was 10), with the 7 inside Steel Line tagged accordingly.
Tests: 7 new (TestSimpleRecipeFlatten) — flat recipes still work,
nested operations surface with correct path label, sub_process
nodes never appear as editor rows, step children of operations
stay hidden, deep-nested sub_processes chain path labels.
Module: fusion_plating 19.0.20.2.0 → 19.0.20.3.0.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Bug surfaced on WO-30043 (2026-05-20): operator walked every step
including a fully closed receiving record, then hit
"Quantity Received is blank — close the receiving record for
SO SO-30043 before completing this job." Receiving WAS closed.
Root cause: the 2026-05-18 cert-creation gate
(fp.job.button_mark_done) blocks on job.qty_received but nothing
populated it. fp.receiving carried the qty on its line records,
fp.job stayed at 0 indefinitely. Two disconnected records on the
same SO.
Fix: when fp.receiving._update_so_receiving_status runs (i.e. on
every state transition — counted / staged / closed / accepted /
resolved), also mirror each line's received_qty onto the matching
fp.job by (sale_order_id + part_catalog_id). Single-part SOs map
1-to-1; multi-part SOs spawn one job per line so the same join
still works.
Two defensive guards in the hook:
- Skip silently when fusion_plating_jobs not installed
(Job = env.get('fp.job') returns None).
- Skip silently when fp.job doesn't yet carry part_catalog_id /
qty_received (test scope, unusual install topology).
Drive-by during cleanup:
- fp_parent_numbered_mixin._fp_assign_parent_name: guard
so.x_fc_parent_number access with field-existence check. The
column lives in fusion_plating_jobs; downstream modules that
inherit the mixin (receiving) but don't depend on jobs were
hitting AttributeError on every fp.receiving.create at test
time. Falls through to the legacy sequence when the column
isn't there.
- fp_receiving_views.xml: legacy carrier_name Char field rendered
as a second carrier row labeled "Legacy Carrier" alongside the
proper x_fc_carrier_id M2O — operators saw two carrier fields
and got confused. Hide the legacy display (data stays in DB for
audit; migration 19.0.3.10.0 already matched it to a real
delivery.carrier).
Migration 19.0.3.19.0/post-migrate.py backfills qty_received from
closed receiving lines for any job stuck at 0 — fixes WO-30043
and two sibling jobs on entech.
Modules: fusion_plating 19.0.20.2.0, fusion_plating_receiving
19.0.3.19.0, fusion_plating_jobs 19.0.10.15.0.
All 19 tests green (TestCarrierFields 6, TestQtyReceivedPropagation 5
new, TestReceivingGate 8). Direct verification on entech: WO-30043
qty_received = 1, mark_done succeeds, delivery + cert auto-created.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Switched the section title from group string= (which Odoo was rendering
as a left-side column label) to a real <separator/>, so the heading
sits above the radio and the options use the full form width.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Reverts the label shortening and instead sets col=1 on the radio group
so the group's inner layout is a single column. With the full wizard
width available, the full labels fit on one line each.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The group title already says "How were pages 11 & 12 provided?", so the
radio labels don't need to repeat "Pages 11 & 12". Shortened to:
"Inside the original application" / "Separate file" / "Sign remotely".
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Bumps fusion_claims version to bust the asset bundle cache after the
Application Received wizard refactor.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Three-mode radio at the top of the Application Received wizard. The
Signed Pages 11 & 12 group is only shown in Separate mode; the remote
sign banner/button is only shown in Remote mode. Adds a read-only
'Detected pages' indicator next to the uploaded original PDF.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Adds intake_mode (bundled / separate / remote) so staff can mark
applications received with a single bundled PDF, the existing
separate-pages-file flow, or a pending remote signature. Folds in
content-based PDF validation, a friendlier status-gate message,
and a page-count helper for the original application.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The signed-pages verification step on case close now treats the bundled
flag as 'pages present', matching the ready-for-submission gate and the
audit trail.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Both the has_documents indicator and the action_confirm missing-items
gate now read x_fc_has_signed_pages_11_12, so orders with pages 11 & 12
bundled inside the original PDF can move to Ready for Submission without
a separate signed-pages file.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
x_fc_trail_has_signed_pages now reads x_fc_has_signed_pages_11_12, so
the trail correctly shows complete when pages 11 & 12 are bundled inside
the original application.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
New boolean on sale.order tracks whether pages 11 & 12 are bundled
inside the original application PDF. Computed helper
x_fc_has_signed_pages_11_12 ORs bundled flag with separate-file and
remote-signing presence so downstream gates can read one field.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Seven-task TDD implementation plan for the design at
2026-05-19-adp-application-received-bundled-pages-design.md. Adds the
bundled-flag + computed gate to sale.order, updates downstream gates
(ready-for-submission, case-close, audit trail), rewrites the
Application Received wizard with a three-mode radio, and bumps the
module version.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Design for refining the Application Received wizard so staff can mark
applications received with a single PDF when pages 11 & 12 are inside
the original application — without losing the existing separate-file
and remote-signing paths.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Operators now drop a .docx or .pdf Fischerscope XDAL 600 export
on the cert form's Thickness Report tab. The wizard parses the
readings, calibration std, operator + date metadata, and the
embedded microscope image, then shows them for review before
recording on fp.certificate.
Operator Wizard Certificate
─────────────────────────────────────────────────────────────
Click "Upload Parse .docx / - thickness_reading_ids
Thickness .pdf → written (3 rows)
Report" Show 3 readings - x_fc_local_thickness
Pick file + metadata _pdf attached (original
Click Parse Click Save file)
- microscope image as
ir.attachment on cert
- chatter post
─────────────────────────────────────────────────────────────
When parse can't find readings (unrecognised format), wizard falls
through to manual state — operator can still save, file lands on
the cert as-is for the existing CoC page-2 merge logic.
Closes the gap in the S19 enforcement: x_fc_send_thickness_report
customers blocked at action_issue until the file is on file. Now
they have a parseable upload UX, not just a bare Binary field.
Architecture
- fischerscope_parser.py: pure-Python lib, branches on extension,
python-docx + PyPDF2 already on entech (no new deps). Regex
extraction returns {readings, metadata, image, errors}.
- fp.thickness.upload.wizard: TransientModel with upload/review/
manual states. Lazy-imports parser at action_parse time to dodge
Python 3.11 partial-init relative-import error.
- 27 tests (TestFischerscopeParser 9 + TestThicknessUploadWizard 8
+ the rehoused TestActionIssueGates 10) — all green on entech.
Same metadata copies onto every reading row, microscope image
attaches once at cert level (decisions 2026-05-19).
Drive-by fixes uncovered while running tests on entech:
- fp.certificate.action_issue: guard rec.company_id access with
field-existence check. Lazy-fill-signer branch crashed when
certified_by_id was unset on certs that don't carry a company_id
field. Pre-existing bug that never fired in production because
jobs auto-fill certified_by_id before reaching this branch.
- test_action_issue_gates: set x_fc_send_thickness_report=False on
the test partner. Field defaults to True so every cert in this
class hit the thickness gate; tests were never able to verify
the other gates in isolation.
- Tests directory missing test_action_issue_gates.py on entech.
Synced; turns out the 2026-05-18 "changes" commit added the file
locally but the deploy script never copied tests/.
Module: fusion_plating_certificates 19.0.6.4.0 → 19.0.7.0.0
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The helper set step_kind_id on fp.job.step when fp.step.kind model
exists, but step_kind_id field doesn't actually exist on fp.job.step
in deployed shape — both test_start_skips_contract_review and
test_finish_skips_contract_review erred with
ValueError: Invalid field 'step_kind_id' in 'fp.job.step'
Per CLAUDE.md rule 18, _fp_is_contract_review_step() matches step
name case-insensitive against 'contract review' or 'qa-005'. The
test only needs to trigger that detection — set name='Contract
Review' on the CR branch and let the receiving gate's existing
exemption fire.
All 8 TestReceivingGate tests now pass on entech.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
@@ -77,6 +77,7 @@ Odoo content-hashes the compiled bundle URL (`/web/assets/<hash>/...`). When CSS
## Cursor-Managed Modules
- **fusion_clock** is currently being modified in Cursor — always read files fresh before editing, don't assume you know the current state
- **fusion_repairs** — status and deferred work: [`fusion_repairs/cloud.md`](fusion_repairs/cloud.md) (bundles 1–11 shipped at `19.0.2.2.4`; not production-deployed)
3. **Backend OWL**: Use standalone `rpc()` from `@web/core/network/rpc`. NOT `useService("rpc")`. `static props = []` not `{}`.
4. **HTTP routes**: `type="jsonrpc"` — NOT `type="json"` (deprecated).
5. **res.config.settings**: Only boolean/integer/float/char/selection/many2one/datetime. NO Date fields.
6. **res.groups**: NO `users` field, NO `category_id` field.
**`config_parameter=` Boolean fields don't round-trip `False` as a string.** Odoo's `set_values()` calls `IrConfigParameter.set_param(key, value)`, and `set_param` deletes the row when `value` is falsy (False / None / empty). So writing `False` to a Boolean config field means the param no longer exists in `ir_config_parameter`; a subsequent `get_param(key)` returns the *default* (Python `False`), not `'False'`. Test like `self.assertFalse(ICP.get_param('...'))` — never `assertEqual(..., 'False')`. (Integer/Float/Char go through `repr(value)` / strip, so they DO persist as strings — `'90'`, `'0'`, etc.) Source: `odoo/addons/base/models/res_config.py::set_values` and `ir_config_parameter.py::set_param`.
6. **res.groups**: NO `users` field, NO `category_id` field. **The Odoo 19 replacement for `category_id` is `res.groups.privilege`.** To make a module's groups appear as application-access dropdowns on the user form (Settings → Users → *Application Accesses*) instead of only in developer mode: define an `ir.module.category`, a `res.groups.privilege` (with `category_id` → that category), and set each group's `privilege_id` → that privilege. Groups under one privilege that form an `implied_ids` chain render as a single role dropdown; a standalone group in its own privilege renders as a separate row under the same category header. Verified in `fusion_clock/security/security.xml`; mirrors `fusion_plating`/`fusion_tasks`.
**res.users**: field was renamed `groups_id` → `group_ids` (also `all_group_ids` for implied). The plural form is gone; using `groups_id` raises `ValueError: Invalid field 'groups_id' in 'res.users'`.
**`ir.ui.view`**: same rename — view-level visibility gating uses `group_ids`, not `groups_id`. A record like `<field name="groups_id" eval="[(4, ref('base.group_system'))]"/>` on an `ir.ui.view` raises `ValueError: Invalid field 'groups_id' in 'ir.ui.view'` at module install. (The XML *attribute* `groups="base.group_system"` on form elements like `<page>`, `<button>`, `<field>` is unrelated and still works.)
**`ir.rule` `groups` field is additive, not restrictive.** A rule with `groups=[some_group]` applies ONLY to users in that group — it does NOT restrict non-members. So `domain_force=[(1,'=',1)]` + `groups=[base.group_system]` does NOT mean "only admins see rows"; it means "admins see all rows (and the rule is silent on everyone else)". Non-admins are gated by the ACL (`ir.model.access.csv`), not the rule. To truly restrict by group at the rule layer, pair a global rule (`groups=[]`, `domain_force=[(0,'=',1)]` = block-all baseline) with a group-scoped allow rule. Default to letting the ACL do the gating; use rules for row-level filters that ACLs cannot express.
7. **Search views**: NO `group expand="0"` syntax.
8. **SCSS imports**: `@import "./partial"` is FORBIDDEN in Odoo 19 custom SCSS. It prints a warning and silently falls back to the old cached bundle. Register every SCSS file (including `_partial.scss` tokens) as a separate entry in `web.assets_backend`. Put tokens first; Odoo concatenates bundle files so SCSS variables/mixins from the first file are visible to every later file.
9. **SQL constraints & indexes**: Odoo 19 dropped `_sql_constraints = [(name, def, msg), ...]` and the `init()`/raw-SQL pattern. Both still parse but only emit a warning and are silently ignored. Use declarative class attributes instead:
```python
_check_qty_positive = models.Constraint('CHECK (qty > 0)', 'Quantity must be positive.')
The attribute name after the leading underscore becomes the SQL object name suffix (`{table}_{suffix}`). `models.Index` accepts `DESC`, `WHERE` predicates, and `USING btree (...)`. Sources: `odoo/orm/model_classes.py` (warns at registry build), `odoo/orm/table_objects.py` (Constraint + Index classes).
10. **`res.users._login` is an instance method in Odoo 19**, not a classmethod as in earlier versions. Signature is `def _login(self, credential, user_agent_env)` — there is no `db` parameter. Override it like any normal instance method (`super()._login(credential, user_agent_env)`). When called via `authenticate()` on an empty recordset, `self` carries the right env. Older recipes that build a separate `api.Environment` from `odoo.modules.registry.Registry(db)` no longer apply. Source: `odoo/addons/base/models/res_users.py:760`.
11. **Inherited `ir.ui.view` records cannot have `groups`/`group_ids` on the record itself.** Odoo 19 raises `ParseError: Inherited view cannot have 'groups' defined on the record. Use 'groups' attributes inside the view definition` at install time. Move the gate to the inner XML nodes — every `<button>`, `<page>`, `<field>`, `<xpath>`, `<group>` etc. supports a `groups="base.group_system"` attribute. For an inherited form with a smart button + admin tab, put `groups=` on the button and the page individually; leave the `<record model="ir.ui.view">` clean.
12. **`mail.template` QWeb/inline_template `ctx` IS `self.env.context`** — not a nested dict you can pass. `MailRenderMixin._render_eval_context()` sets `ctx = self.env.context`, so `ctx.get('foo')` in subject/body resolves to `env.context.get('foo')`. To pass dynamic data to a template, spread keys directly into the context: `tmpl.with_context(**my_data).send_mail(res_id, ...)`. Calling `tmpl.with_context(ctx=my_data)` puts the dict at `env.context['ctx']`, and the template's `ctx.get('foo')` becomes `env.context.get('foo')` → `None` (looks like a silent rendering bug — subject ends up blank).
13. **`ir.cron` dropped `numbercall`** in Odoo 19. Old recipes set `<field name="numbercall">-1</field>` for "run forever"; that now raises `ValueError: Invalid field 'numbercall' in 'ir.cron'` at install time. Just omit the field — recurring crons keep running as long as `active=True`. Source: `odoo/addons/base/models/ir_cron.py` field list.
14. **`cr.commit()` / `cr.rollback()` raise AssertionError inside `TransactionCase`** — they are NOT silent no-ops in Odoo 19. The test cursor explicitly refuses both ("Cannot commit or rollback a cursor from inside a test, this will lead to a broken cursor when trying to rollback the test. Please rollback to a specific savepoint instead..."). For cron/worker code that needs per-row isolation so one bad row doesn't roll back the whole batch, use `with self.env.cr.savepoint(): ...` inside the loop instead of `cr.commit()`. Savepoints work in both prod (under the outer cron transaction) and tests (under the outer test transaction). The cron transaction commits the whole batch when the method returns; in tests everything rolls back cleanly. Source: `odoo/sql_db.py::TestCursor.commit` and `Cursor.savepoint()`.
15. **There is NO `sale.subscription` model in Odoo 19** (Enterprise `sale_subscription`). A subscription is a **`sale.order`** with `is_subscription=True`, `plan_id` → **`sale.subscription.plan`** (the recurrence), plus `subscription_state` / `next_invoice_date` / `recurring_monthly`. Any Many2one or relation that targets "a subscription" must point at `sale.order` (filter `domain=[('is_subscription','=',True)]`) — **not** `sale.subscription`, which does not exist and fails at install. The surviving `sale.subscription.*` records are only the plan + wizards/reports (`sale.subscription.plan`, `sale.subscription.report`, `sale.subscription.change.customer.wizard`, `sale.subscription.close.reason.wizard`). Verified on live `nexamain` (odoo-nexa, 19.0): `SELECT model FROM ir_model WHERE model LIKE 'sale.subscription%'`.
16. **Renaming a module's technical name needs a DB rename, not just a folder rename.** The technical name is baked into the database: `ir_module_module.name`, every external ID in `ir_model_data.module`, each view's `ir_ui_view.key` prefix, and the `ir_module_module_dependency.name` rows of every module that depends on it. Rename only the folder + in-code references and Odoo treats the new name as a fresh uninstalled module — installing it **duplicates** groups/templates/menus and **orphans** all existing data. On every DB that already has it installed, run an in-place SQL rename (the 4 tables above) **before** `-u <newname>`; a fresh DB needs nothing. Reference script + full rationale: [`fusion_portal/rename_module.sql`](fusion_portal/rename_module.sql) (written for the `fusion_authorizer_portal` → `fusion_portal` rename). Also update cross-module `depends`, `inherit_id="<old>.view"`, `t-call`, `env.ref('<old>.xmlid')`, asset paths (`<old>/static/...`), and `from odoo.addons.<old>... import`.
## Card Styling — Copy Odoo's Kanban Pattern
Don't rely on `var(--bs-border-color)` or `var(--bs-body-bg)` for card surfaces — they drift between themes/addons and often render **invisible**. Odoo's own kanban (`.o_kanban_record`) uses **explicit hex** values:
@@ -75,14 +94,40 @@ Odoo content-hashes the compiled bundle URL (`/web/assets/<hash>/...`). When CSS
- Canadian English for all user-facing text
- Currency: `$` sign with Monetary fields + currency_id
## Cursor-Managed Modules
- **fusion_clock** is currently being modified in Cursor — always read files fresh before editing, don't assume you know the current state
## Module-Specific Notes
- **fusion_clock** — developed in **Claude Code** (no longer Cursor; no concurrent-editing conflicts). Changed a lot recently (NFC kiosk: tap-to-clock, enrollment + program-from-unknown-tap, manager page, sounds, screen lock, guided profile-photo capture, faster animations). Still read files fresh before editing rather than assuming the layout. Live on entech (`odoo-entech` / LXC 111 on `pve-worker5`).
- **fusion_repairs** — read [`fusion_repairs/cloud.md`](fusion_repairs/cloud.md) before feature work. **Version `19.0.2.2.4`.** Bundles 1–11 shipped in repo (intake, portals, dashboard, pricing, flowcharts, parts/PO). **Not production-deployed** to Westin as of 2026-05-27. Local: `docker exec odoo-modsdev-app odoo -d fusion-dev -u fusion_repairs --stop-after-init`. Outstanding: RingCentral SMS, C2 history sidebar UI, office follow-up crons (config keys only), `tests/`, more flowchart content, sales-rep dashboard tile in `fusion_portal`.
- **Running module tests requires ephemeral ports.** The dev container's main Odoo process holds 8069 and 8072; a `docker exec ... odoo --test-enable` will die with `Address already in use` unless you also pass `--http-port=0 --gevent-port=0`. This is because Odoo 19 forces `http_spawn()` when `--test-enable` is set, even when `--no-http` is passed. Canonical test invocation:
- **`fusion_centralize_billing` tests run on odoo-trial (VM 316).** Local dev is Community and cannot install this module. Use `bash scripts/fcb_test_on_trial.sh` from the repo root. The script uses `--http-port 8070` to avoid the port 8069 conflict with the live odoo-trial-app container. Pass = `FCB_EXIT=0`. Takes ~1-2 min.
- **Python deps not bundled with `odoo:19` image:** `user_agents` (used by `fusion_login_audit`), and likely others. Install ephemerally with `docker exec -u 0 odoo-modsdev-app pip install <pkg> --break-system-packages`. The install is LOST when the container is recreated (e.g. `docker compose up -d` after a compose edit). When this happens, the symptom is `ModuleNotFoundError` deep in the auth or report code. Re-run the pip install. A persistent fix would be a custom Dockerfile or a startup hook on the compose service — not done yet.
- Test before deploying. Edit existing files — don't create unnecessary new ones.
## PDF Preview — Prefer fusion_pdf_preview Over Downloads/New-Tab
When a Python action opens an attachment, route it through `fusion_pdf_preview` instead of returning `ir.actions.act_url` with `download=true` or `target=new`. The preview dialog gives operators preview + print + download in one place and writes an audit log; non-PDF attachments fall back to the legacy download path automatically.
The drop-in replacement is the new helper on `ir.attachment`:
The helper auto-detects mimetype: PDFs go to the dialog, everything else (ZPL, CSV, XML, images) stays on download. So a callsite that today serves CSV today and a PDF tomorrow doesn't need a code change — same call, different routing.
If you need to invoke the client action directly (rare — only when you don't have a recordset handy), the tag is `fusion_pdf_preview.open_attachment` and the params are `{attachment_id, title, model_name, record_ids, report_name}`. See `fusion_pdf_preview/static/src/js/open_attachment_action.js`.
Existing reports (`ir.actions.report` of type `qweb-pdf`) are intercepted automatically by `fusion_pdf_preview/static/src/js/pdf_preview.js`; the helper above is for the *other* pattern — attachments opened by custom buttons.
## Supabase Knowledge Base
Before starting unfamiliar work, check Supabase for context:
| **2d — Dual-run reconciliation** | `fusion.billing.reconciliation` diffs Odoo-computed vs NexaCloud-actual per customer/period for ≥ 1 cycle before any flip | Safety gate before flipping real billing |
The core engine already built the *receiving* side (`/usage`, webhook engine, charge math).
#2 is about **connecting NexaCloud to it and proving the numbers match before flipping.**
## Decision pending (resume here)
We were in the `superpowers:brainstorming` flow for #2 and stopped at: **which slice to
start with?**
- **(recommended) 2a — Mapping + importer** — lowest risk, foundation for everything else.
- 2d — Reconciliation first (front-load the trust mechanism).
> **For agentic workers:** REQUIRED SUB-SKILL: Use superpowers:subagent-driven-development (recommended) or superpowers:executing-plans to implement this plan task-by-task. Steps use checkbox (`- [ ]`) syntax for tracking.
**Goal:** Attach real customer identity to every helpdesk ticket and give client-deployment staff an in-app ticket inbox (read replies + follow up without leaving their Odoo), while external customers use the native Enterprise portal + magic link.
**Architecture:** Keystone = pass `partner_email`/`partner_name`/`x_fc_client_label` in the ticket-create payload; native helpdesk then creates the partner + subscribes the follower. Client module (`fusion_helpdesk`) gains read/reply RPC endpoints + a tabbed dialog + unread badge, all scoped server-side by the logged-in user's identity. Central module (`fusion_helpdesk_central`) adds the `x_fc_client_label` field + a branded acknowledgement email.
**Tech Stack:** Odoo 19 (Enterprise on central, Community on client deployments), Python 3.11, OWL 2, XML-RPC client→central, `helpdesk` (Enterprise), `portal.mixin`, `mail.thread.cc`.
**Testability note:**`fusion_helpdesk` depends only on base/web/mail → installable + testable on local Community (`odoo-modsdev-app`, DB `modsdev`). Pure logic (scope-domain, message filtering, vals builder, unread math) is extracted into `fusion_helpdesk/utils.py` and unit-tested with no live remote. `fusion_helpdesk_central` depends on `helpdesk` (Enterprise) → install/test on the deploy target (odoo-nexa) or odoo-trial.
---
## File Structure
**`fusion_helpdesk` (client)**
-`utils.py`*(new)* — pure helpers: `build_scope_domain`, `is_public_message`, `build_ticket_vals`, `compute_unread_count`. No Odoo env needed → trivially unit-testable.
-`controllers/main.py`*(modify)* — keystone payload in `submit()`; new endpoints `my_tickets`, `ticket_detail`, `ticket_reply`, `unread_count`; a mockable `_rpc(model, method, args, kw)` seam.
- [ ]**Step 1:** Inherit the helpdesk ticket list + search to add `x_fc_client_label` (column `optional="show"`, search field + a group-by). Use `group_ids` not `groups_id` if gating (none needed here).
> NOTE at execution: verify the exact `inherit_id` external IDs by reading the live views (`helpdesk.helpdesk_ticket_view_tree`, `helpdesk.helpdesk_tickets_view_search`) on odoo-nexa — names differ across versions. Adjust before install.
**Files:** Create `fusion_helpdesk/security/fusion_helpdesk_groups.xml`; Modify `__manifest__.py` (add to `data`, FIRST so the group exists before ACLs reference it if needed)
- [ ]**Step 2:** Implement endpoints (all `type='jsonrpc'`, `auth='user'`). `my_tickets` builds the scoped domain via `build_scope_domain`, `search_read` fields `[id, name, stage_id, write_date]`, plus a per-ticket latest public support message id (read `message_ids` or a dedicated query), then computes `has_unread` via the seen map. `ticket_detail` re-resolves the ticket through the scoped domain (reject if absent), reads public messages only (filter via `is_public_message` using each message's subtype internal flag fetched from central), and calls `_mark_seen`. `unread_count` returns `compute_unread_count(...)`.
> Execution detail: fetch message subtype "internal" flag from central by reading `mail.message` fields `[author_id, date, body, message_type, subtype_id]` and resolving `subtype_id.internal` via a second read or by filtering `message_type='comment'` + excluding notes. Confirm the cleanest field set against the live `mail.message` model during execution.
- [ ]**Step 1:** Endpoint `/fusion_helpdesk/ticket/<int:ticket_id>/reply`, `auth='user'`. Re-resolve ticket via scoped domain (reject if not in scope). Resolve author partner on central by the replier's email (find-or-create via `res.partner` search/create through bot, or pass `author_id` resolved from `partner_email`). Post:
- [ ]**Step 1:** Add to state: `tab:'new'|'list'|'thread'`, `tickets:[]`, `loadingList`, `current:{id,subject,messages,canReply}`, `replyBody`, `replyEmail` (default from a new `/fusion_helpdesk/whoami` or seeded via session user email — read `user.email` via `useService('user')`/`session`), `scope:'mine'|'all'`, `isAdmin`.
- [ ]**Step 2:** Methods: `openList()` → rpc `/fusion_helpdesk/my_tickets` (with `scope`); `openTicket(id)` → rpc detail, switch to thread, refresh list badge; `sendReply()` → rpc reply then reload thread; `setScope()` (admin toggle). Add confirmed **Your email** input on the New tab bound to `state.replyEmail`, passed as `reply_email` in submit payload.
- [ ]**Step 3:** Template: a tab header (New | My Tickets); New pane = existing form + email field; List pane = table (ref, subject, stage chip, unread dot) + admin Mine/All toggle; Thread pane = messages (author, date, body, attachments) + reply box + Back. Use `Markup`-safe rendering: render message bodies with `t-out` (OWL) since central returns sanitized HTML.
- [ ]**Step 4:** SCSS for tabs/list/thread (follow Odoo kanban hex pattern + dark-mode `$o-webclient-color-scheme` branch per CLAUDE.md).
- [ ]**Step 1:** On setup, call `/fusion_helpdesk/unread_count`; store `state.unread`. Poll on an interval (e.g. 120s) and on dialog close. Show a badge bubble when `unread > 0`.
- [ ]**Step 2:** Badge markup over the icon. **Step 3: Commit**
- [ ]**Step 1:**`mail.template` on `helpdesk.ticket` with subject "We received your request [{{ object.ticket_ref }}]" and a body using the company email layout + a prominent button to `{{ object.get_base_url() }}{{ object.access_url }}` (magic link). Canadian English.
- [ ]**Step 2:** Send on create via a create-override (central inherit), gated:
iftmplandt.partner_emailandt.x_fc_client_label:# in-app channel only → avoid double-ack with native web form
tmpl.send_mail(t.id,force_send=False)
returntickets
```
> Decision: gate on `x_fc_client_label` so only in-app-channel tickets get OUR ack; external web/email customers rely on native confirmation (verify native behavior during deploy; widen the gate if native sends nothing).
- [ ]**Step 3:** Register template data in manifest; **Step 4: Commit**
---
## Phase 6 — Review, fix, deploy, smoke test
### Task 13: Code review + fix
- [ ] Run the code-review skill / pr-review-toolkit `code-reviewer` + `silent-failure-hunter` over the diff. Fix HIGH/MEDIUM findings. Re-run client tests locally. Commit fixes.
### Task 14: Deploy + test central on odoo-nexa
- [ ] Copy/confirm `fusion_helpdesk_central` source is visible to odoo-nexa (`/opt/odoo/custom-addons`).
- [ ] Run module tests on nexa: `-u fusion_helpdesk_central --test-enable --test-tags /fusion_helpdesk_central --stop-after-init` (ephemeral http port). Fix failures.
- [ ] From entech: file ONE test ticket via the dialog (or simulate the controller path).
- [ ] On nexa: confirm the new ticket has `partner_id` resolved, `partner_email`/`partner_name`/`x_fc_client_label` set, customer is a follower, ack email queued/sent.
- [ ] Reply as agent on nexa → confirm notification email to the reporter w/ magic link; confirm the entech dialog "My Tickets" shows the ticket + reply and the badge increments.
- [ ] Confirm pre-existing identity-less tickets are untouched (the "lots already submitted" set) and do NOT leak across deployments in the inbox query.
---
## Self-Review (run before execution)
- **Spec coverage:** keystone (T1-3), label field+views (T3-4), scoping (T5,8,9), seen/badge (T6,10,11), admin group (T7), ack email (T12), portal/native (config — verified live, no code), tests (T1,5,6 local + T3 enterprise), deploy+smoke (T14-16). ✓
- **Placeholders:** none — code shown for all Python/XML; JS tasks specify state/methods/markup concretely. JS is manually QA'd (OWL unit tests out of scope).
> **For agentic workers:** REQUIRED SUB-SKILL: Use superpowers:subagent-driven-development (recommended) or superpowers:executing-plans to implement this plan task-by-task. Steps use checkbox (`- [ ]`) syntax for tracking.
**Goal:** Build a one-time, re-runnable, read-only importer that backfills NexaCloud customers/plans/deployments into Odoo as a shadow copy (drafts, no charge) for dual-run reconciliation.
**Architecture:** A `fusion.billing.import.wizard` transient model. `_read_nexacloud_rows()` opens a read-only `psycopg2` connection (DSN from `ir.config_parameter`) and returns plain row dicts — the only code touching NexaCloud. `_import_rows(data, dry_run)` is pure Odoo: it upserts the `nexacloud` service, a `cpu_seconds` metric, Monthly/Yearly recurrences, partners+links (reusing `_resolve_or_create_partner`), a per-plan catalog (product + CPU-overage product + `fusion.billing.charge` with `plan_id` left NULL), and one **draft** shadow `sale.order` per deployment with the flat price set explicitly on the line. Shadow-safety holds by construction: draft + no payment token + charge `plan_id` NULL.
- **Never code Odoo internals from memory** (repo CLAUDE.md rule #1). The uncertain internals (`recurring_invoice`, `is_subscription` on a draft order, `sale.subscription.plan` fields, `price_unit` stickiness, `sale.subscription.plan``billing_period_unit` values) are *verified by the tests themselves* on odoo-trial — when a test fails because an assumption is wrong, fix the source, do not weaken the assertion.
- **Models, not UI:** all logic lives in `_import_rows` / `_do_import` / `_import_*` model methods; the wizard button only calls them. This keeps everything testable under `TransactionCase`.
- **Money:** CAD, prices are `Float`/`Monetary`. CPU overage: `price_per_unit=0.0075`, `unit_batch=3600`.
- **New fields on native models:** `x_fc_*` prefix.
- **Registering tests:** append `from . import test_importer` to `tests/__init__.py` in the task that creates it; commit `__init__.py` alongside so the package always imports.
## Test environment
Tests run on **odoo-trial** (Proxmox VM 316, Odoo 19 Enterprise, db `trial`) — local dev is Community and cannot install this module. One runner:
```bash
bash scripts/fcb_test_on_trial.sh
```
- It re-syncs the module to the sandbox and runs `-u fusion_centralize_billing --test-enable --test-tags /fusion_centralize_billing`.
> **Note:** `partner_by_user` and (Task 3) `plan_ctx_by_id` are **method-local** dicts — never set them as attributes on `self` (Odoo recordsets reject arbitrary attribute assignment). Tasks 3 and 4 add their loops to this same `_do_import` method, so the locals stay in scope.
- [ ]**Step 4: Run it, expect pass**
Run: `bash scripts/fcb_test_on_trial.sh`
Expected: `FCB_EXIT=0`; `TestImporterIdentity` passes. If `country_id.code` assertion fails, fix `_fc_resolve_country` (don't weaken the assertion).
Expected: `FCB_EXIT=0`; both catalog tests pass. If `product.product` rejects `recurring_invoice` or `type='service'`, read the field on odoo-trial and fix the source.
Expected: `FCB_EXIT=0`. If `is_subscription` is False on the draft order, that disproves the design assumption — read `sale_order.py` in `sale_subscription` on odoo-trial and adjust how the subscription is created (e.g. set the field driving `is_subscription`), never weaken the assertion. If `billing_period_unit` rejects `'year'`, read the selection values and fix `_fc_recurrence_plan`.
Expected: `FCB_EXIT=0` — idempotency and dry-run already hold from Tasks 2–4 + the savepoint in `_import_rows`. If the dry-run leaves a `nexacloud` service behind, the savepoint isn't wrapping `_fc_service` — confirm `_do_import` (which creates the service) runs entirely inside the `with self.env.cr.savepoint()` block.
"charges with NULL plan_id must keep the rating cron a no-op")
```
- [ ]**Step 2: Run it, expect pass**
Run: `bash scripts/fcb_test_on_trial.sh`
Expected: `FCB_EXIT=0` — the safety properties hold by construction (draft, no token, NULL plan_id). If `payment.token` is not a valid model name in this build, read the `payment` model names on odoo-trial and use the correct one (don't drop the assertion). If an invoice *is* found, the draft-import guarantee is broken — investigate whether `sale.order.create` auto-invoices, and stop confirming/posting.
Expected: `FCB_EXIT=0` — the per-row `try/except` + `savepoint` already isolates failures. If the whole batch aborts, the `savepoint` is missing around `_import_user` or the broad `except` is too narrow — fix so one bad row never poisons the cursor.
Expected: `FCB_EXIT=0` — `_read_nexacloud_rows` raises `UserError` when the DSN param is empty (implemented in Task 1). If `psycopg2` import fails on odoo-trial, confirm it ships with the image (it does — Odoo depends on it).
Expected: no undefined names (catches the kind of `_norm_email` NameError the helpdesk smoke test missed).
- [ ]**Step 5: Commit (if any fixes)**
```bash
git add -A fusion_centralize_billing/
git commit -m "test(billing): 2a importer full suite green + static checks"
```
---
## Done = 2a importer complete
A NexaCloud backfill produces, idempotently: unified partners + links, a `cpu_seconds` charge catalog (`plan_id` NULL), and one draft shadow `sale.order` per deployment carrying the exact NexaCloud flat price — with zero customer-visible billing in Odoo (no invoice, no token, rating cron a no-op). The `psycopg2` read path is ready; the live run is gated only on the read-only DSN grant.
**Goal:** Ingest NexaCloud's real (Stripe-billed) invoices into Odoo as posted `account.move` customer invoices with reconciled payments + HST, so Odoo is the accounting system of record — all history + ongoing, revenue split by service family, draft-first on the live books.
**Architecture:** A new ingester in `fusion_centralize_billing` mirroring the importer's read/write split: `_read_nexacloud_invoices` (read-only psycopg2 via the existing DSN) → `_ingest_invoices` (pure Odoo: create `account.move` drafts idempotently, map lines to per-family income accounts, derive tax, reconcile Stripe payments) → `_post_ingested` (bulk-post after review). Reuses the `account.link` partner mapping. Native Odoo accounting does the rest.
- **Never code accounting internals from memory** (CLAUDE rule #1). Reference confirmed on trial: `account.move` has `invoice_line_ids`/`invoice_date`/`action_post`; `account.payment.register` exists; `account_type='income'`/`'asset_receivable'` valid; sale taxes are Canadian (find HST 13% by `amount=13` / name). Where a step says "read reference", confirm before relying on it.
- **Models, not UI:** logic in model methods; the wizard only calls them. Testable under `TransactionCase`.
- **New fields on native models:** `x_fc_*`. Declarative `models.Constraint` only.
- Tests run on **odoo-trial** (`bash scripts/fcb_test_on_trial.sh`, full suite, ~1–2 min). Register each new `tests/test_*.py` in `tests/__init__.py` in the same task.
- [ ]**Step 4: run** → PASS. (If `account.account.create` needs more required fields on this build, read `account_account.py` on trial and add them — don't weaken the test.)
- [ ]**Step 4: run** → PASS. (Read reference if no 13% sale tax exists: `docker exec odoo-trial-app ... grep -i hst` the l10n_ca data; on nexamain confirm the HST 13% record from `nexa_coa_setup`.)
Confirm `account.move.create({'move_type':'out_invoice','partner_id':..,'invoice_line_ids':[(0,0,{'name','quantity','price_unit','account_id','tax_ids'})]})` and `move.amount_untaxed/amount_tax/amount_total`.
"""Placeholder until Task 5; defined so post=True doesn't AttributeError."""
returnFalse
```
- [ ]**Step 4: run** → PASS. (If tax computes to 13.00 only when the company/fiscal position allows it, read the tax setup on trial; if `amount_tax` ≠ 13.00, the chosen tax is wrong — fix `_fc_tax_for`, never weaken the assertion.)
- [ ]**Step 4: run** → PASS. (If `payment_state` is `in_payment` rather than `paid`, that's expected when the bank journal isn't reconciled to a statement — accept both, as the assertion does.)
- [ ]**Step 5: commit** — `feat(billing): reconcile Stripe payments so ingested invoices show paid`
Then (separate, gated, NOT in this plan): on nexamain — prune shadow data, **dry-run** the full backfill (review the per-family $ summary + unmatched "Other" lines), ingest **as draft**, you review a sample, **bulk-post**, enable the daily cron.
**Goal:** Compute, per shadow subscription + period, Odoo's would-be charge vs NexaCloud's actual charge and record the delta in `fusion.billing.reconciliation`, so the dual-run can prove parity before any flip.
**Architecture:** A pure `_compute_reconciliation(...)` (testable) + `_reconcile_rows(rows)` (resolves the shadow sub → flat + charge, upserts recon rows) + a read-only `_read_reconciliation_rows()` (psycopg2, integration glue). Triggered from the import wizard + cron. Odoo-only; reads NexaCloud, writes only reconciliation rows.
- [ ]**Step 4: set it in the importer.** In `wizards/import_wizard.py``_import_subscription`, add the plan id to both the `shadow_vals` dict (so re-runs keep it current) :
- [ ]**Step 1: add the reader** in `wizards/import_wizard.py` (reuses the DSN + the same connect/guard pattern as `_read_nexacloud_rows`). Aggregate usage cpu_hours per (subscription, period) and the invoice subtotal per (subscription, period); return rows shaped for `_reconcile_rows`:
The dual-run can be run each cycle (button/cron): it reads NexaCloud usage + invoice subtotals, computes Odoo's would-be charge, and records per-subscription `match`/`delta` rows. Flip happens (manually) once a cycle is all-match.
# Fusion Clock — Province-Aware Automatic Unpaid Break Implementation Plan
> **For agentic workers:** REQUIRED SUB-SKILL: Use superpowers:subagent-driven-development (recommended) or superpowers:executing-plans to implement this plan task-by-task. Steps use checkbox (`- [ ]`) syntax for tracking.
**Goal:** Make the unpaid meal break deduct automatically from worked hours on every path (portal, kiosk, NFC, cron, **and manual backend entry**), using a 2-tier per-province rule table (Ontario: 5h→30min, 10h→+30min), with no duplicated logic.
**Architecture:** A new `fusion.clock.break.rule` table holds the per-province thresholds. `hr.employee._get_fclk_break_rule()` resolves an employee's rule from its company's province (global default fallback). `hr.attendance.x_fclk_break_minutes` becomes a single stored **computed** field — `statutory_break(worked_hours) + Σ penalty_minutes` — that recomputes on every save and replaces the four scattered write sites (controller `_apply_break_deduction`×3 call sites, the auto-clock-out cron, and the penalty code's manual write).
(robocopy exit codes < 8 = success.) **Preflight:** if `K:\Github\odoo-modsdev\addons\fusion_clock` does not exist, the dual-tree setup changed — STOP and confirm the active copy with the user before continuing.
**Container/DB:**`odoo-modsdev-app` / db `modsdev` (per memory `reference_docker_env_names`).
**Canonical commands** (note the ephemeral ports — `--test-enable` forces `http_spawn()` so 8069/8072 collide without them; per repo CLAUDE.md):
**Commit:** only from the git tree (`git -C "K:/Github/Odoo-Modules" ...`). Per memory `feedback_always_push_to_main`, push after each commit on `main`.
---
## File Structure
**Created:**
- `fusion_clock/models/clock_break_rule.py` — the `fusion.clock.break.rule` model + tier engine + constraints.
**Behaviour-change note (intentional, approved by spec §4.3):** today a *late-in* penalty written at clock-in (e.g. +15) is silently swallowed at clock-out because `_apply_break_deduction` does `max(break, current)`. The new compute makes **all** penalty minutes strictly additive (`statutory + Σ penalties`), so a late-in penalty on a long shift is no longer lost. Net hours for such shifts will be correctly lower than before.
<p class="o_view_nocontent_smiling_face">Create a statutory break rule</p>
<p>Define unpaid meal-break thresholds per province/country. Employees inherit
the rule matching their company's province, or the default rule.</p>
</field>
</record>
</odoo>
```
- [ ] **Step 7: Add the menu** — in `fusion_clock/views/clock_menus.xml`, insert after the `menu_fusion_clock_locations_config` menuitem (the Locations config item) and before `menu_fusion_clock_nfc_enrollment`:
```xml
<menuitem id="menu_fusion_clock_break_rules"
name="Break Rules"
parent="menu_fusion_clock_config"
action="action_fusion_clock_break_rule"
sequence="25"
groups="group_fusion_clock_manager"/>
```
- [ ] **Step 8: Wire the manifest** — in `fusion_clock/__manifest__.py`:
**Do NOT bump the version yet** — it stays `19.0.4.0.3` until Task 4, so the
`19.0.4.1.0` migration actually fires in dev (Odoo only runs a version's migration
when the installed version is *lower* than the manifest version).
Add the seed data file after `'data/ir_config_parameter_data.xml',`:
```python
'data/clock_break_rule_data.xml',
```
Add the view file after `'views/clock_schedule_views.xml',`:
```python
'views/clock_break_rule_views.xml',
```
(Data and view files reload on every `-u` regardless of the version number, so the
new model/menu install without a bump. No assets change in this plan, so the bump's
only purpose is the migration trigger — deferred to Task 4.)
Expected: module upgrades cleanly; `test_break_minutes_for_tiers`, `test_second_tier_must_exceed_first`, `test_single_default_enforced` PASS. (Other tests in the class will error until Tasks 2–3 add their dependencies — that's expected if you scoped the run; otherwise the not-yet-added methods simply don't exist yet.)
Expected: FAIL — `AttributeError: 'hr.employee' object has no attribute '_get_fclk_break_rule'`.
- [ ] **Step 3: Implement the resolver** — in `fusion_clock/models/hr_employee.py`, add this method immediately after the `_get_fclk_break_minutes` method (after its `return float(...)` block, before `_get_fclk_scheduled_times`):
```python
def _get_fclk_break_rule(self):
"""Return the statutory break rule for this employee.
Resolution: company's province → matching rule; else the global default
rule; else an empty recordset (caller treats as zero break). Read via
sudo so the portal net-hours compute can resolve it without a direct ACL.
This task is atomic: once the field is computed (no inverse), any remaining `write({'x_fclk_break_minutes': ...})` raises at runtime, so the field conversion and the removal of all four write sites must land together.
Sync, then run the module tests. Expected: the new tests FAIL — e.g. `test_manual_attendance_applies_statutory_break` asserts 30 but gets 0 (no write override exists yet).
- [ ] **Step 3: Convert the field to a stored compute** — in `fusion_clock/models/hr_attendance.py`, replace the field definition:
OLD:
```python
x_fclk_break_minutes = fields.Float(
string='Break (min)',
default=0.0,
tracking=True,
help="Break duration in minutes to deduct from worked hours.",
)
```
NEW:
```python
x_fclk_break_minutes = fields.Float(
string='Break (min)',
compute='_compute_fclk_break_minutes',
store=True,
tracking=True,
help="Unpaid break deducted from worked hours: statutory break (per the "
"employee's province rule, from actual hours worked) plus any penalty "
"minutes. Computed automatically on every save.",
)
```
- [ ] **Step 4: Add the compute method** — in the same file, insert this method immediately before the `_compute_net_hours` method (just above its `@api.depends('worked_hours', 'x_fclk_break_minutes')` decorator):
Expected: no output (clean). If it flags `pytz` as unused in `hr_attendance.py`, that's fine only if no other code uses it — verify before removing the import (the absence/overtime crons still use `pytz`, so leave the import).
- [ ] **Step 9: Run to verify all Task 3 tests pass**
Sync, then run the module tests. Expected: all `test_manual_*`, `test_under_first_threshold_no_break`, `test_penalty_minutes_are_additive`, `test_master_toggle_off_zero_statutory`, `test_open_attendance_zero_break` PASS, and the existing NFC/kiosk/dashboard tests still PASS.
git -C "K:/Github/Odoo-Modules" commit -m "feat(fusion_clock): auto-apply statutory break via one stored compute" -m "x_fclk_break_minutes is now statutory(worked_hours) + penalties, recomputed on every path including manual backend entry. Removes the four duplicated write sites (controller _apply_break_deduction + 3 call sites, auto-clock-out cron, penalty write)." -m "Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>"
help="Automatically deduct the statutory unpaid break from worked hours. Break lengths and thresholds are configured per province under Configuration → Break Rules.">
Sync, then run the module tests. Expected: module upgrades cleanly and the `19.0.4.1.0` migration executes (installed `19.0.4.0.3` < manifest `19.0.4.1.0`; modsdev shows the INFO line, nexa/entech run `log_level=warn`), `test_dead_settings_removed` PASS, full `fusion_clock` suite green.
- [ ] **Step 7: Verify the param is gone and historical rows recomputed** (sanity)
- Attendances → create a manual attendance, check-in 09:00 check-out 15:00 (6h) → **Break = 30**, Net = 5.5h, with no clock action.
- Edit that record's check-out to 19:00 (10h) → **Break = 60**, Net = 9.0h.
- Create a 4h attendance → **Break = 0**.
- Settings → the old "Min. Shift" threshold field is gone; the Auto-Deduct Break help points to Break Rules.
- [ ] **Step 3: Update the module CLAUDE.md** — in `fusion_clock/CLAUDE.md`:
- §4 Model Map: add a row — `fusion.clock.break.rule | models/clock_break_rule.py | Per-province statutory unpaid-break thresholds (2-tier).`
- §5 Clocking Flow: note that the break deduction is no longer a controller step — `x_fclk_break_minutes` is a stored compute (`statutory(worked_hours) + Σ penalties`) that fires on every path including manual backend entry; resolved rule via `hr.employee._get_fclk_break_rule()` (company province → default).
- §13 Gotchas: add — "Unpaid break is computed, not written: never `write({'x_fclk_break_minutes': ...})`; change the province rule (`fusion.clock.break.rule`) or `auto_deduct_break` instead. Penalty minutes are now strictly additive (the old `max()` that swallowed late-in penalties is gone)."
git -C "K:/Github/Odoo-Modules" commit -m "docs(fusion_clock): document province break rules + computed break field" -m "Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>"
git -C "K:/Github/Odoo-Modules" push
```
- [ ] **Step 5: Report** — summarize what changed, the behaviour-change note (penalties now additive), and that live deployment to entech (`odoo-entech`) is a separate step pending user sign-off.
**2. Placeholder scan** — every step has full code/commands; no TBD/TODO/"similar to". ✓
**3. Type/name consistency** — `break_minutes_for`, `_get_fclk_break_rule`, `_compute_fclk_break_minutes`, fields `break1_after_hours/break1_minutes/break2_after_hours/break2_minutes/is_default`, model `fusion.clock.break.rule`, access id `model_fusion_clock_break_rule`, action `action_fusion_clock_break_rule`, menu `menu_fusion_clock_break_rules` — all used identically across tasks. The compute folds `Σ penalty_minutes` (field `penalty_minutes` on `fusion.clock.penalty`, confirmed). ✓
# Accessibility Funding-Source Selector — Implementation Plan
> **For agentic workers:** REQUIRED SUB-SKILL: superpowers:executing-plans (inline) — this is a 3-file change. Steps use `- [ ]` checkboxes.
**Goal:** Let the rep mark an accessibility assessment's funding source (Private / March of Dimes / ODSP / WSIB / Hardship / Insurance / Other) on the web form, so the generated sale order routes to the correct funding pipeline instead of always defaulting to private pay.
**Architecture:** The model (`fusion.accessibility.assessment.x_fc_funding_source`) and the SO routing (`_create_draft_sale_order` → `sale_type_map` → `x_fc_sale_type`) already exist (the "2026-04 portal audit fix"). The only gaps: (1) the form has no funding field, (2) the save controller never reads `funding_source` from the POST, (3) `hardship` is missing from the selectable funding sources. The submit JS already serialises every named form field via `FormData`, so no JS change is needed.
**Verification constraint:**`fusion_portal` depends on Enterprise `knowledge`, so it can NOT be installed on the local Community Docker. Syntax-check with host Python; functional verification is on westin (or a clone): pick "March of Dimes" on a form → the draft SO gets `x_fc_sale_type='march_of_dimes'` and lands in the MOD pipeline.
---
### Task 1: Add Hardship to the funding source + route it
- [ ]**Step 1:** Add `('hardship', 'Hardship Funding')` to the `x_fc_funding_source` selection list (after `'wsib'`).
- [ ]**Step 2:** Add `'hardship': 'hardship',` to `sale_type_map` in `_create_draft_sale_order` (the target `x_fc_sale_type='hardship'` already exists in `fusion_claims``sale_order.py:332`).
- [ ]**Step 3:**`python -m py_compile fusion_portal/models/accessibility_assessment.py` → no error.
- [ ]**Step 4:** Commit.
### Task 2: Add the funding select to the shared client-info form
- [ ]**Step 1:** Add a new row with a `<select name="funding_source">` (options mirror the model selection; `direct_private` pre-selected so existing private behaviour is unchanged) right after the phone/email row, before the card closes.
- [ ]**Step 2:** Validate XML well-formedness (`[xml]` parse).
- [ ]**Step 3:** Commit.
### Task 3: Capture funding_source in the save controller
- [ ]**Step 2:** Deploy to westin (backup → scp the 3 files → `-u fusion_portal` → cache-bust → restart) and confirm: open `/my/accessibility/stairlift/straight`, pick "March of Dimes", complete → the new SO shows `x_fc_sale_type = march_of_dimes` and appears in the MOD pipeline.
In day-to-day operations the office or the client often scans (or emails) the **entire** ADP application as a single PDF — already including signed pages 11 & 12. Today, staff have to manually split pages 11 & 12 out of the bundled PDF and upload them again as a separate file, even though the same signatures are already present in the original PDF.
The wizard must continue to support the existing flows (separate signed-pages file, remote signing via Page 11 signing request), but it should also accept the bundled case without manual splitting.
## Goals
- Allow staff to mark Application Received with **one** PDF when pages 11 & 12 are inside it.
- Preserve the two existing modes (separate file, remote signing).
- Keep downstream audit/case-close checks correct without rewriting every consumer.
- Make the wizard easier to use and slightly safer (real PDF detection, friendlier messages).
## Non-Goals
- PDF page extraction or splitting (explicitly rejected by user — "no split").
- Capturing Page 11 signer identity in the bundled / separate-file modes (existing gap; out of scope).
- Re-architecting the document-attachment model to de-duplicate identical binaries (out of scope).
- Changes to the remote signing wizard or `fusion.page11.sign.request` model.
## High-Level Approach
Add a **single boolean flag** on `sale.order` that records whether pages 11 & 12 are inside the original application PDF. Introduce a **computed helper field** that downstream consumers read instead of `x_fc_signed_pages_11_12` directly. Add a **three-mode radio** at the top of the Application Received wizard.
Minimal blast radius:
- One new boolean, one new computed field on `sale.order`.
- Wizard view + Python rewritten to drive logic off the radio mode.
- Four downstream call sites change which field they read (no logic change).
- Three small complementary fixes folded in (status-gate text, PDF magic-bytes check, page-count indicator).
## Data Model
### `sale.order` — new fields
```python
x_fc_pages_11_12_in_original=fields.Boolean(
string='Pages 11 & 12 in Original Application',
default=False,
tracking=True,
help='True when the original application PDF already contains the signed pages 11 & 12.',
)
x_fc_has_signed_pages_11_12=fields.Boolean(
string='Has Signed Pages 11 & 12',
compute='_compute_has_signed_pages_11_12',
store=True,
help='True if pages 11 & 12 are satisfied — either bundled, uploaded separately, '
`x_fc_trail_has_signed_pages` already exists at [models/sale_order.py:3248](../../fusion_claims/models/sale_order.py:3248). Its compute body changes from `bool(order.x_fc_signed_pages_11_12)` to `order.x_fc_has_signed_pages_11_12`.
### Migration
None. Existing records get `x_fc_pages_11_12_in_original = False` by default; their existing `x_fc_signed_pages_11_12` binary continues to satisfy the new computed gate. Stored compute will populate `x_fc_has_signed_pages_11_12` for legacy rows on first read or recompute.
('bundled','Pages 11 & 12 are INCLUDED in the original application'),
('separate','Pages 11 & 12 are a SEPARATE file'),
('remote','Pages 11 & 12 will be SIGNED REMOTELY'),
],
string='Intake Mode',
required=True,
default='bundled',
)
original_page_count=fields.Integer(
string='Original PDF Page Count',
compute='_compute_original_page_count',
)
```
`signed_pages_11_12` and `signed_pages_filename` keep their current definitions — they're only required in `separate` mode now.
The existing computed fields `has_pending_page11_request` and `has_signed_page11` ([wizard/application_received_wizard.py:44-49](../../fusion_claims/wizard/application_received_wizard.py:44)) **stay** — they drive the "request pending" / "remote signature complete" banners now only shown when `intake_mode == 'remote'`.
### `default_get` — pick an initial mode from existing state
```python
# When re-opening the wizard on an order that already has some data:
Page count is displayed read-only next to the original-application filename once a PDF is loaded. If `pdfrw` fails to parse, show *"(could not read PDF)"* — does not block confirmation.
When `intake_mode == 'bundled'`, any pre-existing `x_fc_signed_pages_11_12` from a prior wizard run is left alone (we don't clear it). The bundled flag plus the existing separate file together are harmless — the computed gate is `OR`.
### PDF magic-bytes check
```python
def_validate_pdf_bytes(self,b64_data,label):
importbase64
ifnotb64_data:
return
try:
head=base64.b64decode(b64_data)[:5]
exceptException:
raiseUserError(f"{label}: could not decode uploaded file.")
ifhead!=b'%PDF-':
raiseUserError(f"{label} must be a PDF file (content check failed).")
```
The existing filename `.pdf` check stays in place as a defence-in-depth `@api.constrains`.
### Chatter message — mode-aware
| Mode | Headline | Detail line |
|---|---|---|
| `bundled` | *Application Received — bundled* | "Pages 11 & 12 included in original PDF" |
| `separate` | *Application Received — separate files* | "Original + separate signed pages uploaded" |
| `remote` | *Application Received — remote signature pending* | "Page 11 sent for remote signature (`N` request(s) outstanding)" where `N` is the count of `page11_sign_request_ids` in state `sent` or `signed`. |
Notes from the wizard, if any, are appended below as today.
## Downstream Consumer Changes
These are mechanical: change which field they read. **No logic changes.**
| File | Line | Old | New |
|---|---|---|---|
| [wizard/ready_for_submission_wizard.py:95](../../fusion_claims/wizard/ready_for_submission_wizard.py:95) | `_compute_field_status` | `bool(order.x_fc_original_application and order.x_fc_signed_pages_11_12)` | `bool(order.x_fc_original_application and order.x_fc_has_signed_pages_11_12)` |
| [wizard/ready_for_submission_wizard.py:148](../../fusion_claims/wizard/ready_for_submission_wizard.py:148) | gate check | `if not order.x_fc_signed_pages_11_12` | `if not order.x_fc_has_signed_pages_11_12` |
The `x_fc_signed_pages_11_12` field stays in the data model. Any download / preview / "open document" button that points at the literal binary stays as-is — bundled-mode orders simply won't have this field populated, and the UI should hide the "Open signed pages" button when the field is empty (it already does — Odoo hides empty binary widgets by default).
## Error / Edge Cases
| Scenario | Behaviour |
|---|---|
| User toggles from `separate` to `bundled` after uploading a separate file | Wizard does not clear the upload field. On confirm, only the original application is written; bundled flag goes to True. The separate-file binary in the wizard is discarded (it was never written). |
| User picks `remote` but has no sent/signed request | Block with the message above; user must click *Request Remote Signature* first. |
| User picks `bundled` but the PDF is short (e.g. 4 pages) | Page-count indicator shows *"(4 pages)"* as a visual hint, but **does not block**. The 14-page ADP form is the norm but the system can't reliably enforce it across form versions. |
| Legacy record without `x_fc_pages_11_12_in_original` set | Defaults to False. As long as `x_fc_signed_pages_11_12` is present, `x_fc_has_signed_pages_11_12` is True — gate still passes. |
| Stored compute not populated for legacy rows | Triggered on first read or via a one-line `_recompute` on module load is **not** required — Odoo computes on first access. If users hit issues, a one-off psql `UPDATE` can be run manually. |
| Remote signing completes after `bundled` mode was used | `_compute_has_signed_pages_11_12` already ORs in `page11_sign_request_ids.state == 'signed'` — harmless overlap; trail stays correct. |
| Uploaded file is not really a PDF (wrong content) | Magic-byte check raises a UserError; record is not changed. |
## Testing
### Unit tests — wizard (`tests/test_application_received_wizard.py`, new)
- Reload browser with cache clear (per CLAUDE.md asset-bundle-cache rule).
- No production deploy steps unique to this change.
## Open Questions (none blocking implementation)
- Should bundled-mode capture Page 11 signer identity (signer name, relationship) the way the remote flow does? Currently neither bundled nor separate-file modes do — existing gap, deferred.
- Should the bundled-mode chatter automatically attach a one-line note like *"Operator confirms pages 11 & 12 are within the original application"* with the user's name? The default chatter post already records the user. Leaving as-is.
A spot audit of user `info@gsafinancialconsulting.com` ("GSA Accounting", uid 63) revealed Odoo's built-in login tracking is effectively unusable for compliance:
-`res.users.log` rows are pruned by the daily `_gc_user_logs` cron — only the most recent login per user survives. For GSA Accounting the entire history collapsed to a single row at `2026-04-22 20:24 EDT`.
-`/var/log/odoo` on the production VM is empty because Odoo is configured at `log_level=warn` with stdout-only logging; INFO-level auth lines aren't captured anywhere.
- The container's json log is 444 KB and rotates frequently — nothing about the user remains.
- The existing `network_logger` module records outbound HTTP traffic from Odoo (uid=1 always), not user activity.
Result: today there is **no durable record** of who logged in, when, from where, or how often. A user with `base.group_system` + Technical Features and no 2FA — like GSA Accounting — could be active for months without any reconstructable trail.
This module closes that gap with a dedicated audit table that survives Odoo's GC, captures successful and failed authentications, surfaces results in the user form, and alerts admins on suspicious failure bursts.
## Goals
1.**Durable audit trail** of every password-authenticated login (success and failure) on `westin-v19`.
2.**Per-user visibility** for Settings admins via a tab + smart button on `res.users`.
3.**Failure-burst alerting** to admins on a configurable consecutive-failure threshold.
4.**Geo-enrichment** of IPs out-of-band so authentication latency is unaffected.
5.**Zero risk to the auth path** — an audit-write failure must never block a real login.
## Non-goals (v1)
- Logging every HTTP request / page view (explicitly de-scoped during brainstorming).
- Logging session resume events from auth cookies.
- API-key authentication (`credential['type'] == 'apikey'`) — bypasses `_check_credentials`. Documented as a known gap; addressable in a follow-up.
- OAuth / SSO logins — no OAuth provider configured on westin-v19.
- Self-service "view my own login activity" for end users — visibility is admin-only.
- Auto-disabling users on failed logins — flagged as a self-service DoS vector during brainstorming.
- 30-day local cache retention_days Login Audit menus
- Settings page section
```
The auth-path hooks are synchronous (must run inside the request). Geolocation, alerting, and retention are out-of-band so they cannot affect login latency.
-`application=False` (it's a technical addon, not a top-level app)
**Dependencies (Python):** none new. Uses the `user_agents` library already shipped with Odoo. Geolocation calls `http://ip-api.com/json/<ip>` via the standard `requests` library (no API key required, 45 req/min free tier).
**Field naming:** new fields on existing models (`res.users`, `res.config.settings`) use the `x_fc_*` prefix per project CLAUDE.md. The new `fusion.login.audit` model uses unprefixed field names.
## Data model
### `fusion.login.audit` (new model, table `fusion_login_audit`)
| Field | Type | Notes |
|---|---|---|
| `user_id` | Many2one(`res.users`, `ondelete='set null'`) | Null if attempted login didn't match any user |
| `attempted_login` | Char(255), indexed | Always set — even on unknown-user failures |
| `x_fc_last_login_ip` | Char(45), compute, store=True | Surfaces last source IP in the form header |
The `store=True` computes are triggered by the create on `fusion.login.audit` (via `@api.depends('x_fc_login_audit_ids.event_time', 'x_fc_login_audit_ids.result')`).
### `res.config.settings` additions
Booleans / integers only (per CLAUDE.md — no Date fields on settings):
Each is backed by an `ir.config_parameter` (`fusion_login_audit.retention_days`, etc.) so changes from the Settings page persist.
### Multi-company
`fusion.login.audit` is intentionally **company-agnostic**. Logins happen before any company context is established; synthesizing one would either break the unknown-user case or require a "system company" placeholder. Settings admins see all rows globally.
## Capture flow
### Successful login (`_update_last_login`)
```python
def_update_last_login(self):
result=super()._update_last_login()
try:
self._fc_record_login_event(result='success')
exceptException:
_logger.exception("fusion_login_audit: failed to record success row for %s",self.login)
returnresult
```
Called by Odoo only after the credential check has passed. Super() runs first so Odoo's own bookkeeping is unaffected.
### Failed login on known user (`_check_credentials`)
_logger.exception("fusion_login_audit: failed to record/alert failure")
raise
```
TOTP failures (from `auth_totp`) also raise `AccessDenied` and are caught here. Distinguish via `credential.get('type') == 'totp'` to set `failure_reason='2fa_failed'`.
### Failed login on unknown user (`_login` classmethod)
_logger.exception("fusion_login_audit: failed to record unknown-user failure")
raise
```
Without this override, unknown-user attempts never reach `_check_credentials` and would silently disappear from the audit. The classmethod sets `user_id=None` and stores the attempted login string.
vals['geo_lookup_state']='internal'# distinct from private_ip; cron skips both
returnvals
```
### Write semantics
- All writes use `self.env['fusion.login.audit'].sudo().create(vals)` — low-privilege users can still generate their own audit rows despite the read-only record rule.
-`mail_create_nolog=True` context to avoid chatter noise.
- The password value is **never** present in `vals` and is hard-stripped from any `credential` dict before logging. A regression test asserts this.
## Async geolocation cron (`cron_geo_enrich`)
**Schedule:** every 5 minutes, `numbercall=-1`, `priority=10`.
**Worker logic:**
1. Select 100 oldest rows where `geo_lookup_state='pending'`.
2. For each row:
- **Private-IP shortcut:** if `ip_address` is in `10.0.0.0/8`, `172.16.0.0/12`, `192.168.0.0/16`, `127.0.0.0/8`, `::1`, or `fe80::/10` → set `geo_lookup_state='private_ip'`, `country_code='--'`, `city='Private network'`.
- **Cache check:** look for any prior row with the same `ip_address` and `country_code IS NOT NULL` and `event_time > now() - interval '30 days'`. If found, copy `country_code` / `country_name` / `city` / `geo_state` / `ip_hostname` locally; set state `done`. No external call.
- **Reverse DNS:** `socket.gethostbyaddr(ip)` with `socket.setdefaulttimeout(1.5)`.
- **HTTP lookup:** `requests.get('http://ip-api.com/json/' + ip, params={'fields': 'status,country,countryCode,regionName,city'}, timeout=3, headers={'User-Agent': 'Odoo-FusionLoginAudit/19.0'})`. The call passes through `network_logger` automatically.
- On `status='success'` → fill fields, set state `done`.
- On HTTP error, timeout, or `status='fail'` → set state `failed` (no retry).
3.`self.env.cr.commit()` after each row so one bad IP cannot roll back the batch.
4.**Rate limit defense:** if the response header `X-Rl` is `'0'`, break early and leave remaining rows as `pending` for the next run.
**Privacy:** the only outbound data is the IP itself. No user identifiers, no Odoo URL, no headers beyond `User-Agent: Odoo-FusionLoginAudit/19.0`. All outbound calls are auditable in `network_logger`.
## UI surfaces
### `res.users` form view
- **Smart button** in the button box, gated `groups="base.group_system"`:
```
┌──────────────┐
│ 🔑 N Logins │
└──────────────┘
```
Click → opens `fusion.login.audit` list view filtered to this user (`domain=[('user_id', '=', active_id)]`).
No new group created. Read is bound to existing `base.group_system`. Rationale: brainstorming decision was "Settings admins only" — reusing the existing group avoids an extra checkbox to manage.
### Model access (`ir.model.access.csv`)
| Group | Read | Write | Create | Unlink |
|---|---|---|---|---|
| `base.group_system` | ✓ | ✗ | ✗ | ✗ |
**No write/create/unlink for any group via the UI.** Audit rows are only written via `sudo()` from inside the auth hooks. An audit log admins can mutate is not an audit log.
### Record rule
Single global rule on `fusion.login.audit`: read for `base.group_system` only. The user-form one2many is additionally gated at the view level via `groups="base.group_system"` (not via a more permissive record rule) so non-admins have no read path even if they craft a custom view.
### Field-level
- `failure_reason` stores a category, never the attempted password.
- `_fc_build_event_vals` strips `credential['password']` before any logging or row construction.
- The `credential` dict is never persisted.
- Regression test: no field on `fusion.login.audit` ever contains a known-test-password string.
## Retention
**Cron `cron_retention_gc`** — daily at 03:00 UTC, `numbercall=-1`:
```python
days = int(self.env['ir.config_parameter'].sudo().get_param(
Uses `unlink()` rather than raw `DELETE` so any ORM side effects fire. Expected DB load on `westin-v19`: 27 users × ~2 logins/day × 365 days ≈ 20k rows steady state — trivial for Postgres.
## Failed-login alert
**Mail template** in `data/mail_template_data.xml`:
- **Subject:** `[Login Audit] {threshold} failed login attempts for {attempted_login}`
- **Body:** simple HTML table of the last N failure rows for that `attempted_login` — timestamp, IP, country, user-agent summary.
- **Recipients:** all users in `base.group_system` with a non-empty `email`.
- **Send path:** `mail.mail` queue with `auto_delete=True` so the auth response isn't blocked.
**Cooldown:** 60 min per `attempted_login`, enforced via an `ir.config_parameter` keyed by `fusion_login_audit.last_alert:{attempted_login}` storing the last-send timestamp. Prevents a sustained attack from flooding admin inboxes.
**Kill-switch:** if `x_fc_login_audit_alert_enabled = False`, no alerts are sent regardless of threshold.
## Edge cases
| Case | Behavior |
|---|---|
| `request` is None (XML-RPC, internal auth from cron) | Row written with `ip_address='internal'`, `user_agent_raw='<no-request>'`, `geo_lookup_state='internal'` (cron skips) |
| Audit insert errors on a hot DB | Login still succeeds — every auth-path hook is wrapped in `try/except Exception: _logger.exception(...)` |
| User deleted while audit rows remain | `ondelete='set null'` preserves history; `attempted_login` keeps the readable identifier |
| Password reset / `auth_signup` | The reset itself generates no login event; the subsequent login does — matches expectation |
| API key authentication | **Out of scope v1** (bypasses `_check_credentials`); documented |
| OAuth / SSO | Out of scope v1; no provider configured on westin-v19 |
| Portal user (`share=True`) | Logged the same way; smart button remains admin-visible |
| Two requests racing on the same private IP | Each writes its own row; geo cache is best-effort, not transactional |
| `proxy_mode = False` in `odoo.conf` | `remote_addr` will be the reverse-proxy IP — known limitation, fixable by setting `proxy_mode = True` (out of scope) |
## Testing
### `tests/test_login_audit.py` (TransactionCase)
1. Successful login writes a row with `result='success'` and resolved `user_id`.
2. Bad password writes `result='failure'` with `failure_reason='bad_password'` and re-raises `AccessDenied`.
3. Unknown user writes `result='failure'` with `failure_reason='unknown_user'`, `user_id=None`, non-null `attempted_login`.
4. No field on the written row contains the attempted password (regression).
5. Geo cron: pending row gets enriched from local cache when same IP exists within 30 days (no HTTP call made).
6. Retention cron: rows older than `retention_days` are deleted; newer survive.
7. Alert email: 5 failures in 15 min queues exactly one `mail.mail`; a 6th failure within cooldown queues zero.
8. `database` field is populated from `self.env.cr.dbname`.
9. Audit-write exception inside `_update_last_login` does not block the login.
### `tests/test_security.py` (HttpCase)
1. Non-admin user gets `AccessError` on direct `search(fusion.login.audit)`.
2. Non-admin sees the user form view without the smart button or "Login Activity" tab (XML node hidden by `groups`).
3. Settings admin sees both.
## Deployment notes
- **Local install:** copy module to `K:\Github\Odoo-Modules\fusion_login_audit\` (bind-mounted into `odoo-modsdev-app` container). Update via:
- **Icon:** copy `C:\Users\gsing\Downloads\fusion logs.png` to `K:\Github\Odoo-Modules\fusion_login_audit\static\description\icon.png`.
- **Verify `proxy_mode = True`** in `/opt/odoo/odoo.conf` on odoo-westin before relying on `ip_address` accuracy — otherwise `remote_addr` will be the reverse-proxy IP rather than the real client. Confirmed out of scope for this module, but flag for the operator.
- **Verify outbound to `ip-api.com:80`** is reachable from the odoo-westin VM (Tailscale/firewall) — if blocked, `geo_lookup_state` will simply be `failed` and the rest of the module is unaffected.
## Success criteria
- Logging in as any user creates exactly one `fusion.login.audit` row with `result='success'` and the correct IP/UA.
- Failed login attempts create exactly one row with `result='failure'` and the correct `failure_reason`.
- Unknown-user attempts create a row with `user_id=None` and the typed login string in `attempted_login`.
- The smart button on `res.users` shows the lifetime count and opens the filtered list.
- The "Login Activity" tab shows the last 30 events with correct color coding.
- After 5 failures from the same login string within 15 minutes, exactly one alert email arrives in the inbox of every Settings admin with an `email` set.
- The geo cron populates `country_code`, `city`, `ip_hostname` for public IPs within 10 minutes of the login.
- The retention cron, set to 1 day for a test, deletes rows older than 24 hours and leaves newer ones.
- Model `fusion.helpdesk.ticket.seen`: per-user read tracking for the badge.
- `res.config.settings`: (existing) — no new config required beyond what exists.
**`fusion_helpdesk_central` (central) — small additions**
- `helpdesk.ticket` inherit: `x_fc_client_label` field + backend list/search exposure.
- `mail.template`: branded acknowledgement on ticket create (with the magic-link CTA).
- Data/doc: confirm the "Customer Care" team portal config (already correct on live — assert via comment or
light data, don't fight existing config).
---
## 6. Surface A — In-app embedded inbox (detail)
### 6.1 Controller endpoints
All `type='jsonrpc'`, `auth='user'`. **Identity is always derived server-side from `request.env.user`** —
never from request parameters. All remote calls go through the existing bot XML-RPC layer.
| Route | Returns | Notes |
|---|---|---|
| `POST /fusion_helpdesk/submit` *(modified)* | `{ok, ticket_id, ticket_url}` | Adds `x_fc_client_label` + `partner_name`; the confirmed form email is sent as `partner_email` (param may be named `reply_email`, but it maps straight to `partner_email`). |
`fusion_pdf_preview`. So Odoo already does subscriptions, recurring invoicing, full
accounting/GL, Stripe, HST taxes, customer portal, credit notes, and self-serve
checkout.
**The only capability Lago has that Odoo lacks natively is usage-based metered
billing** (billable metrics → aggregation → quota/overage charges). That, plus the
integration surface, is all we build.
Prior decision on record (Supabase `fusionapps.decisions`): Lago was deployed as the
centralizer for NexaDesk + NexaCloud. This design **supersedes** that — the billing
brain moves into the Odoo Enterprise already owned and operated.
## 3. Decisions locked in this session
1.**Odoo fully replaces Lago.** Build a metered-billing engine inside `fusion_centralize_billing`; decommission Lago VM 318 at the end.
2.**One unified customer, separate invoice per service.** One `res.partner` per real client; each service bills on its own subscription/cycle. No cross-product invoice merging.
3.**Apps drive; Odoo is the billing system of record.** Each app keeps its own signup, provisioning, and entitlement enforcement, and calls Odoo's billing API (the same way it calls Lago today). Odoo invoices, charges Stripe, and emits webhooks back.
4.**Odoo owns the billing catalog; apps own entitlements.** Odoo is SoR for products, prices, recurrence, metric rate/quota/overage, taxes — keyed by a stable `plan_code`. Apps enforce feature limits (max_chatbots, CPU quota, API rate-limit) against the same code.
5.**Pilot = NexaCloud, phased dual-run cutover** (one product at a time, parallel run + reconciliation before flip).
6.**Aggregate-push usage ingestion.** Apps push periodic pre-aggregated counters; Odoo stores rollups and feeds native `sale.subscription` metered lines. No raw-event firehose into Odoo.
- **Parent:** Sub-project #2 (NexaCloud adapter + dual-run reconciliation). This spec covers **chunk 2a only** — the read-only importer/backfill. 2b (usage wiring), 2c (control loop), 2d (reconciliation) are separate specs.
- **Depends on:** the core engine (sub-project #1, on `main` at `d770c0c3`): service registry, `_resolve_or_create_partner`, `fusion.billing.charge._compute_billable`, `fusion.billing.usage`, the inbound API, the webhook engine.
## 1. Goal
Backfill the **existing** NexaCloud customers, plans, and deployments into Odoo so the
central billing engine has a complete shadow copy to run dual-run reconciliation (2d)
against. The importer is a **one-time, re-runnable** migration — *not* a continuous sync.
New NexaCloud signups after the cutover already flow through the live inbound API built in
sub-project #1.
The importer must be **safe by construction**: while NexaCloud is still the live biller,
nothing the importer creates in Odoo may charge, post, or email a customer.
## 2. Decisions locked in brainstorming (2026-05-27)
1.**Per-deployment granularity.** NexaCloud's own `subscriptions` table carries
`deployment_id` + `plan_id`, so the natural mapping is **one Odoo subscription
`sale.order` per deployment**. (Confirms spec §15 Q2.)
2.**Billing model = flat plan price + metered overage.** Customers pay a fixed
monthly/yearly plan price PLUS per-unit charges for usage above the plan's quota.
(Confirms the original §6 quota+overage assumption.)
3.**CPU metric standardized to `cpu_seconds`.** The NexaCloud plan quota
(`plans.cpu_seconds_quota`) is already in seconds, so it maps to `charge.included_quota`
with no conversion. NexaCloud's CPU rate ($0.0075/core-hour) maps to
- **Module:** `fusion_centralize_billing` (Odoo 19 Enterprise; build/test on odoo-trial, run on `nexamain`)
- **Supersedes (for NexaCloud):** the metered-billing direction (recompute charges from a CPU-seconds model). The dual-run proved that model captures ~6% of reality.
## 1. Why this exists (the pivot)
The dual-run reconciliation (2026-05-27) showed **94% of NexaCloud's revenue is billed
outside** the per-deployment/CPU-metered model the engine was built for:
| NexaCloud invoices | count | total |
|---|---|---|
| NOT linked to a `subscriptions` row (Hosting services, add-ons) | 22 | **$2,881.08** |
| Linked to a `subscriptions` row (what the metered importer reads) | 7 | **$180.79** |
NexaCloud bills via **Stripe** — service invoices (Odoo ERP Hosting / WordPress Hosting
~$214.50/mo), **add-ons** (Daily Backup, WhatsApp, Forms Builder, White Label), and
**Stripe proration** ("Remaining time on …"). That billing already works. **Re-implementing
Stripe's proration + add-on logic in Odoo is the wrong move.** Instead, Odoo **ingests
NexaCloud's actual invoices** and becomes the single **accounting system of record**
(posted invoices + reconciled payments + HST), while NexaCloud/Stripe keep doing the billing.
- **Status:** Design (proceeding straight to build — approach determined by parent spec §10)
- **Module:** `fusion_centralize_billing` (Odoo 19 Enterprise; tested on odoo-trial)
- **Parent:** Sub-project #2. Depends on **2a** (the importer creates the shadow subscriptions + the `cpu_seconds` charge catalog this reconciles against).
Some in-app feature requests and bug reports require sign-off from a real decision-maker at the client (the "owner" — the person paying the bill, not just an Odoo Manager-by-permission). Today this happens out-of-band via WhatsApp or phone, leaving no record on the ticket and forcing Gurpreet to remember who said what to whom.
We need a structured way to loop the client's owner in on tickets that need approval, on-demand from the central support side, with a low-friction approve/reject flow for the owner and a transcript of the decision living on the ticket itself.
## Goals
- Central support (Gurpreet on nexa) decides *which* tickets need approval — never automatic.
- Owner approves or rejects with **one click** from their email, no login required.
- The approval decision is **publicly visible** on the ticket (per existing chatter / inbox plumbing) — both the originating employee and central support see who approved or rejected and any optional comment.
- Owner contact lives in **entech settings** (source of truth) and stays automatically fresh on nexa via piggyback on every ticket submission.
- An **AI summary** of the ticket goes in the approval email so the owner can decide in 30 seconds without reading the whole thread.
- **Single-shot reminder** if no response in N days.
- **Bulk engagement** when multiple requests need the same owner's sign-off in one batch.
- **Reporting dashboard** so Gurpreet can spot stuck approvals at a glance.
## Non-goals
- Manager-tier approvals (rejected during brainstorming — "manager" by Odoo permission ≠ business-authority owner; only owner needed).
- SLAs / hard deadlines on owner response.
- Multi-step approval chains (one owner, one decision).
- Owner-facing mobile app or portal beyond the approve / reject confirmation page — email + magic link is the entire UX.
- Auto-progressing the ticket stage on approval — Gurpreet still manually completes the work.
## Architecture
### Module split
| Module | Role | Touches |
|---|---|---|
| `fusion_helpdesk` (entech, client) | Lets the client configure their owner contact; sends contacts upstream on every ticket | 2 ICP settings, settings view, `/fusion_helpdesk/submit` payload |
| `fusion_helpdesk_central` (nexa) | Owns the engagement flow end-to-end: storage, wizard, email, public portal, reminder cron, dashboard | New wizard model, ticket fields, mail template, public controllers, OpenAI client, reporting views |
### Data model
#### Entech (`fusion_helpdesk`)
Two new `ir.config_parameter` keys exposed in **Settings → Fusion Helpdesk → Owner Approval**:
-`fusion_helpdesk.owner_email` — Char
-`fusion_helpdesk.owner_name` — Char
`controllers/main.py::submit` piggybacks both keys on every ticket payload (alongside the existing identity keys). Both are optional — leaving them blank disables the Engage button on central for that client.
#### Central (`fusion_helpdesk_central`)
Extend existing `fusion.helpdesk.client.key` (one row per client deployment):
| Field | Type | Purpose |
|---|---|---|
| `owner_email` | Char | Current owner contact for this client. Upserted on every incoming ticket from the submit payload. |
| `owner_name` | Char | Display name for greeting / chatter attribution. |
Extend `helpdesk.ticket`:
| Field | Type | Purpose |
|---|---|---|
| `x_fc_engagement_state` | Selection (`none`/`pending`/`approved`/`rejected`) | Drives kanban badge + state pill on form. Default `none`. |
| `x_fc_engagement_email` | Char | Snapshot of owner email reached for *this* engagement. Survives later edits to `client_key.owner_email`. |
| `x_fc_engagement_name` | Char | Snapshot of owner name. |
| `x_fc_engagement_token` | Char (UUID4) | Single-use token in the magic link. Cleared on confirm. |
| `x_fc_engagement_sent_at` | Datetime | When the engagement email was first queued. |
| `x_fc_engagement_reminded_at` | Datetime, nullable | When the single reminder went out. Set by cron. |
| `x_fc_engagement_decided_at` | Datetime, nullable | When state transitioned to `approved`/`rejected`. Drives turnaround metric. |
| `x_fc_ai_summary` | Text | The brief used in the email; editable in the wizard before send; read-only after. |
1. Support opens the ticket → clicks **`Request Owner Approval`** (header button; only rendered when `x_fc_client_label` is set and `client_key.owner_email` is configured).
- **AI Summary** textarea — auto-populated on `default_get` via one OpenAI call against `{ticket.name + html2plaintext(ticket.description) + each public chatter message}`. Editable.
- **Personal note** textarea — Gurpreet's own one-liner that prepends the email body.
- Read-only display of `owner_email` / `owner_name` resolved from `client_key`.
- Mail template `mail_template_engagement` rendered → queued (`mail.mail`, `auto_delete=True`)
- Wizard closes
4. Owner receives email → reads → clicks **`Approve`** or **`Reject`** (two big buttons, each a `https://erp.nexasystems.ca/fusion_helpdesk/engagement/<token>/<decision>` URL).
5. Public controller resolves the token → renders a small standalone QWeb page (not the heavy portal layout):
- The chatter message propagates to the employee's My Tickets thread via the existing `_public_messages` filter, satisfying the "Fully visible" UX choice.
- Gurpreet receives the standard Odoo follower notification.
7. Support sees the state pill flip from amber `⏳ Awaiting approval from Kris` to green `✓ Approved by Kris`, then progresses the ticket as normal.
### Re-engagement
If Gurpreet clicks **`Request Owner Approval`** on a ticket that's already `pending` / `approved` / `rejected`, the wizard opens normally; on send it overwrites the token, snapshot fields, summary, `sent_at`, and clears `reminded_at` and `decided_at`. State resets to `pending`. Old chatter messages from prior engagements stay as audit history. Old tokens are immediately dead (the token field has changed).
### Token security
UUID4 is 122 bits of entropy — sufficient against guessing. Tokens are single-use (cleared on confirm). No date-based expiry in v1 — keep it simple; if abuse appears, add a 14-day `engagement_sent_at` cutoff in the controller.
## AI summary (OpenAI integration)
- Model: `gpt-4o-mini` (configurable via ICP). ~$0.15/1M input tokens; one call per Engage click. ~$0.01/month at 10 engagements/week.
- Transport: `urllib.request` against `https://api.openai.com/v1/chat/completions` — no new pip dependency.
- Timeout: 15 seconds. On failure → summary field renders empty + soft banner "AI summary unavailable — write a quick brief manually." Wizard remains usable.
- HTML stripping: `odoo.tools.mail.html2plaintext()` (built-in).
- Token cap: assembled prompt truncated to 8000 characters (well below context window, bounds cost on tickets with 50+ messages).
- Prompt is a Python constant (`fusion_helpdesk_central/utils.py::SUMMARY_PROMPT`) so it's editable in one place without UI churn. See Engagement Wizard for prompt text.
- **Privacy**: ticket description + chatter goes to OpenAI. Document in client onboarding. Empty API key disables the auto-fill but keeps the wizard working with a manual summary.
- **Reminder subject** (when wizard's `is_reminder=True`, set by cron): `Reminder: still waiting on your approval — "{{ ticket.name }}"`
- **Body**: branded HTML matching the existing ack template style; greeting uses `engagement_name`; includes personal note, summary, full description + chatter in a `<details>` collapsible, two big approve/reject buttons.
### Public approval portal
Routes (both `auth='public'`, `csrf=False`):
- `GET /fusion_helpdesk/engagement/<token>/<string:decision>` — renders the confirmation page (or "no longer valid" page if token / state invalid). `decision` is validated against `('approve', 'reject')`.
- `POST /fusion_helpdesk/engagement/<token>/<string:decision>` — accepts optional `comment` form field, performs the state transition + chatter post, renders a "Thanks — your decision is recorded" page.
Token resolution helper `_resolve_engagement(token, decision)` returns the ticket or raises a friendly error if anything's off. Used by both GET and POST.
## Bulk engagement
Server action on `helpdesk.ticket` list view: **`Request Owner Approval (bulk)`**.
### Validation (hard errors)
- All selected tickets share the same `x_fc_client_label` — otherwise: "Cannot bulk-engage tickets across different deployments."
- All selected tickets have `engagement_state in ('none', 'rejected')` — otherwise: "{n} of the selected tickets already have a pending or approved engagement. Engage them individually."
- `client_key.owner_email` is configured for the deployment — otherwise the standard tooltip error.
### Wizard
Same `fusion.helpdesk.engagement.wizard` model gains a `ticket_ids` Many2many to `helpdesk.ticket` (single-ticket mode keeps using `ticket_id`; the wizard checks which is set and branches). Per-ticket AI summaries generated **in parallel** via `concurrent.futures.ThreadPoolExecutor(max_workers=5)` with a 30-second overall timeout. Each per-ticket summary is editable in its own row in the wizard view via a child transient model `fusion.helpdesk.engagement.wizard.line` (fields: `wizard_id`, `ticket_id`, `ai_summary`).
### Email
A single combined email with one card per ticket. Each card has its own `[Approve][Reject]` buttons, each pointing at that ticket's unique token. Owner can decide per-ticket, ignore some, come back to the same email later (links stay live until clicked or re-engaged).
### Layout (rendered HTML)
```
Hi Kris,
5 requests from ENTECH need your sign-off. Each can be approved or
rejected independently — clicking a button on one card only acts on
that card.
──── Request 1 of 5 ──────────────────────────────
"Drag and drop steps"
• <summary bullets>
[✓ Approve] [✗ Reject]
──── Request 2 of 5 ──────────────────────────────
...
```
## Reminder cron
`ir.cron`, daily at 09:00, sudo:
```python
N = int(ICP.get_param('fusion_helpdesk_central.engagement_reminder_days') or 3)
4. **Kanban (default group by state)**: at-a-glance count per state
Filters: by client, by state, by date range. Canned filter "Pending > 7 days" highlights stuck approvals.
No new model; everything is derived from `helpdesk.ticket`. The stored computed field `x_fc_engagement_turnaround_hours` makes the pivot fast on large datasets.
## UI changes
### Helpdesk ticket form (nexa)
- New header button **`Request Owner Approval`** (visible iff `x_fc_client_label` set AND `client_key.owner_email` set; tooltip on disabled state explains why)
- `approved` → green `✓ Approved by {{ engagement_name }}`
- `rejected` → red `✗ Rejected by {{ engagement_name }}`
- New collapsible group **`Owner Engagement`** showing `ai_summary` (read-only after send), `engagement_email`, `engagement_name`, `engagement_sent_at`, `engagement_reminded_at`, `engagement_decided_at`, `engagement_turnaround_hours`
### Helpdesk ticket kanban (nexa)
Amber corner dot when `engagement_state == 'pending'` — surfaces blockers in the kanban view without opening each card.
### Entech settings UI
New section **Owner Approval** under existing Fusion Helpdesk group:
- `Owner email` text input
- `Owner name` text input
- Help text: "Used when Nexa Systems support requests approval for a feature or bug fix that needs sign-off. Leave blank if your deployment doesn't require approvals."
## Edge cases
| Case | Behaviour |
|---|---|
| Owner contact not configured on entech | `Request Owner Approval` button disabled, tooltip: "Owner contact not configured for this client. Ask them to fill it in under Settings → Fusion Helpdesk." |
| Token reused after first click | Friendly "This approval link has already been used or is no longer valid" page with a `mailto:support@nexasystems.ca` link. |
| Owner gets re-engaged | New token replaces old; old immediately invalid. State resets to `pending`. Old chatter is preserved. `reminded_at` / `decided_at` cleared. |
| OpenAI down / no API key | Wizard opens with empty summary + soft banner; you type your own brief, send normally. |
| Owner replies to the email instead of clicking | Mail gateway treats it as a regular comment (existing flow). State stays `pending` until they click a magic link. |
| Employee files a follow-up while owner is deciding | Reply lands in chatter normally; owner sees it next time they reload, but their engagement is tied to the snapshot AI summary (intentional — owner judges a stable artifact). |
| Bulk action selects tickets across clients | Hard error before wizard opens. |
| Bulk action selects tickets that already have pending engagements | Hard error specifying the count of disallowed tickets. |
| Approved ticket needs to be "reversed" | No undo button. Re-engage with a fresh wizard → new summary → re-send. Audit chain stays in chatter. |
## Tests
Pure helpers in `fusion_helpdesk_central/utils.py` (new file):
- Wizard `action_send` writes all expected fields, queues mail, returns close action
- Re-engagement clears the old token + decided_at + reminded_at, resets state to `pending`
- Public controller rejects invalid / used / wrong-decision tokens with friendly error
- Public controller `POST` confirms decision, posts chatter, writes state
- State transitions are correctly one-way (approved → approved is no-op, approved → re-engaged → pending works)
- Bulk wizard rejects mixed-client selection
- Bulk wizard rejects already-pending tickets in selection
- Reminder cron only acts on rows past cutoff and not already reminded
- Computed `turnaround_hours` matches expected delta after decision
OpenAI is mocked in tests — no live API calls in CI.
## Versions
- `fusion_helpdesk` → bump to `19.0.2.0.0` (minor feature, new settings)
- `fusion_helpdesk_central` → bump to `19.0.2.0.0` (major feature, multiple new fields + wizard + controllers + cron + reporting)
## Deployment order
1. Deploy `fusion_helpdesk_central` first (it owns the storage, the wizard, the email template, the public routes, the cron, the reporting). It can sit dormant — no Engage button is reachable until `client_key.owner_email` is populated.
2. Deploy `fusion_helpdesk` second (adds the entech settings + payload piggyback). First ticket filed after this deploy populates `client_key.owner_email` on central.
3. Backfill: for any client that already has owner contact info known to Gurpreet (e.g., entech → kris@enplating.ca), edit the `client_key` row directly on nexa via the existing config UI. Or simply wait — the next ticket from that client will populate it.
**Module:**`fusion_portal` (depends on `fusion_claims`, `fusion_tasks`); live on `odoo-westin` (DB `westin-v19`)
**Status:** Draft for review
**Author:** Brainstormed with Gurpreet (Fusion / Westin Healthcare)
---
## 1. Problem & goals
A sales rep visits a client's home **with an occupational therapist (OT) and the client present for only 30–45 minutes**, and the OT's time is the scarcest resource. In that window the team often does more than one assessment — a wheelchair (ADP) plus, opportunistically, accessibility products the rep spots (a ramp at the front steps, a stair lift inside, a tub cutout, a patient lift for transfers). Today each assessment is a **separate, standalone web form** that re-collects the client's details and creates its own sale order, and the front-end forms give the rep **no way to mark a case's funding source** — so March-of-Dimes work silently defaults to private pay and never reaches the MOD pipeline.
**Goals**
1.**One visit, many assessments, entered once.** Bundle every assessment from one home visit; capture the client + funding details a single time.
2.**Measurement-first.** Capture measurements while the OT is present; defer client/health-card data to after they leave; let the OT sign the ADP application on the spot.
3.**Add as you go.** The rep adds an assessment/product the instant they spot it — repeatable, with a location tag (Front / Back / Inside).
4.**Route by funding workflow.** On completion the visit emits **one sale order per funding workflow** (ADP, March of Dimes, ODSP, WSIB, private, …) — never one combined SO, and never a separate SO per item within the same funding.
5.**Let the rep set funding at assessment time** (the real MOD "tracking" gap).
6.**ADP multi-device** with valid-combination rules, including a new **mobility scooter** type and a **home-accessibility hard rule** for power mobility that feeds the accessibility upsell.
**Non-goals (v1):** voice/dictated entry; rebuilding the measurement math; a new MOD/ADP claim model (the pipelines already exist — we reuse them).
---
## 2. Current state (verified against source)
- **Two assessment models, already two separate SO lineages.** `fusion.assessment` (ADP: rollator/wheelchair/powerchair) and `fusion.accessibility.assessment` (the 7 lift/mod types) each have their own `_create_draft_sale_order` (`assessment.py:587`, `accessibility_assessment.py:751`), their own `x_fc_sale_type`, and their own state machine — ADP's 24-state `x_fc_adp_application_status` vs MOD's 16-state `x_fc_mod_status`. Each guards against a second SO (`accessibility_assessment.py:503-511`). SO back-links are **scalar** Many2one: `assessment_id`, `accessibility_assessment_id` (`fusion_portal/models/sale_order.py:37,48`).
- **SOs are born with no order lines.** Specs become a **chatter HTML note** (`_format_assessment_html_table`, `accessibility_assessment.py:815`); a human prices the draft afterward. **No per-type product mapping exists.**
- **Funding is modelled but not on the measurement forms.** `x_fc_funding_source` (required, default `direct_private`) on the accessibility model — values `march_of_dimes`, `odsp`, `wsib`, `insurance`, `direct_private`, `other` (`accessibility_assessment.py:71-87`) — is present on the public booking form but **absent from all 7 measurement forms**, so they default to private. Canonical billing type `sale.order.x_fc_sale_type` (`fusion_claims/models/sale_order.py:320`) carries the full set incl. `adp`, `adp_odsp`, `march_of_dimes`, etc.
- **MOD tracking already exists** as `x_fc_mod_status` (16 states) + ~60 `x_fc_mod_*` fields (HVMP reference #, vendor code, drawings, PCA, POD, approved/payment amounts, dated audit trail) + MOD views + ~7 wizards + ~40 MOD/ODSP stage emails (`fusion_claims/models/sale_order.py:438,877`). An accessibility assessment funded `march_of_dimes` already lands its SO in this pipeline at `need_to_schedule`. **The gap is purely that the rep can't choose `march_of_dimes` on the form.**
- **Emails** are mostly Python-built via the shared `fusion.email.builder.mixin._email_build` (`fusion_tasks/models/email_builder_mixin.py:8`), gated by `ir.config_parameter``fusion_claims.enable_email_notifications`. Completion email fires from inside `_create_draft_sale_order` (`assessment.py:847`; `accessibility_assessment.py:624`). Stage emails (`_adp_send_stage_email`, `_mod_email_build`, `_odsp_email_build`) are keyed off the SO's funding type + status, so **they keep working per-SO unchanged**.
- **Known bug:** backend ADP `action_complete()` sends the authorizer **two** completion emails (template pair at `assessment.py:494` + inline report via `:847`). Must consolidate before fanning out across a visit.
---
## 3. The design
### 3.1 The Visit aggregate (only net-new model)
`fusion.assessment.visit` — the hub for one home visit.
- **Links to its assessments:** `adp_assessment_ids` (One2many → `fusion.assessment`) and `accessibility_assessment_ids` (One2many → `fusion.accessibility.assessment`). Each assessment gains `visit_id`.
- **Links to its sale orders:** `sale_order_ids` (One2many → `sale.order`) — one per funding workflow it produced.
- On the SO side, add `visit_id`. Each assessment already carries `sale_order_id` (Many2one — `accessibility_assessment.py:153`, `assessment.py:422`), so several same-funding assessments can already point at one SO; the redundant **scalar**`assessment_id` / `accessibility_assessment_id` on the SO (`fusion_portal/models/sale_order.py:37,48`) become **One2many** (or are dropped in favour of the `sale_order_id` reverse) so an SO no longer assumes a single source assessment.
Client info moves to the Visit as the single source of truth; the per-assessment `client_name`-required gate is relaxed (the model keeps the field for back-compat / standalone use but the Visit flow fills it from `partner_id`).
### 3.2 Add-as-you-go workspace (portal UX)
A portal "visit workspace" (reps are portal users, tablet-first):
- Always-present **"+ Add"** → pick a type + location tag (Front / Back / Inside / custom) → drop **straight into the existing measurement form** for that type. No client paperwork required to start.
- Each added assessment is a **card** showing type, location, status (To measure / Measured / Signed), and — once priced — its amount.
- **Measurement-first:** the forms render with client fields hidden/optional; a **deferred "Client + funding" step** is completed after the OT leaves and is shared by every item.
- The **OT signs the ADP application (Page 11)** inline on the wheelchair/ADP item, on-site, independent of client demographics (reuse `portal_assessment_express` Page-11 section + signature pad).
- Mockups (for reference, in repo `docs/mockups/` if committed): `fusion_portal_new_approach_mockup.html`.
### 3.3 Multi-instance + location tags
Any type can be added **more than once**, each its own assessment record with a **location label** ("Main stairs", "Basement", "Front porch"). Two stair lifts = two assessment records (→ two lines on the same funding SO; see §3.6). A **"Same as the previous"** action copies shared options so the rep only re-enters the differing measurements.
### 3.4 Per-item funding selector — the MOD gap fix
Expose `x_fc_funding_source` on **each accessibility assessment** in the flow: **Private Pay / March of Dimes / ODSP / WSIB / Hardship / Insurance / Other**. This one field drives the existing `sale_type_map` → `x_fc_sale_type` → correct pipeline (MOD 16-state tracker, ODSP, hardship, …). Defaults to the previous item's funding so an all-MOD visit isn't re-picked each time. **ADP/wheelchair items are fixed to ADP** (no picker). This is the minimal change that closes the "can't mark a case as March of Dimes" gap — no new tracking model.
> **Patient lift** is an accessibility/equipment item that uses this same picker — funded by March of Dimes, **ODSP**, or **Hardship** (e.g. Toronto residents), so its funding is chosen per case, not fixed.
> **`sale_type_map` gap:** `x_fc_funding_source` currently lacks `hardship` while `x_fc_sale_type` already has it (`sale_order.py:320`) — add `hardship` to the picker + a `sale_type_map` entry (`accessibility_assessment.py:771`), and review the map so every offered funding routes to a real `x_fc_sale_type`.
> **MOD funding cap** applies to MOD items — see Resolved decision 1 (§4).
**Multi-device ADP order.** Today one ADP device per order; the visit allows a **valid combination** of ADP devices for one client, all landing on the **one ADP SO**. Each ADP device is an item; the combination check runs across the visit's ADP items.
Rule in words: **at most one "seated-mobility" device** {manual wheelchair, power wheelchair, scooter}, **optionally one walker/rollator alongside, no duplicates.** Enforced when adding/saving an ADP device.
**Scooter (new ADP type) fields:**`client_weight` (exists), scooter type, **maximum travel range**, and the home-accessibility check (below). Gets its own measurement section in the ADP form, mirroring the rollator/wheelchair/powerchair sections.
**Power-mobility home-accessibility hard rule.** For **scooter and power wheelchair**, a required check: *"Is the home accessible enough for the device to be used **inside and outside** the home independently — no lifting, not left outside/in the garage?"* ADP will not fund power mobility a home can't accommodate. If the answer is **No**, the visit **flags an accessibility need** and prompts the rep to add an accessibility item (ramp / porch lift, typically March of Dimes) to remediate. This is the explicit bridge between the ADP power-mobility item and the accessibility/MOD upsell.
> **The power-wheelchair form is already well-optimized — do NOT change its fields.** The *only* addition there is this home-accessibility warning. The new **scooter** type gets its own section (fields above); the manual-wheelchair and rollator sections are unchanged.
### 3.6 Funding-workflow grouping → one SO per workflow
On visit completion, group its assessments by **funding workflow** (`x_fc_sale_type`) and create **one SO per group**:
- All `march_of_dimes` items (stair lift + porch lift + tub cutout, or two stair lifts) → **one MOD SO, multiple lines** (funding permitting).
- All ADP devices (the valid combination) → **one ADP SO**.
- Private / ODSP / WSIB / insurance → their own SO each.
- A separate SO appears **only when the case type changes**, never per-item within a funding.
Refactor the two per-model `_create_draft_sale_order` routines into a **shared, group-aware builder** that takes a set of same-funding assessments and produces one SO, branching on funding type to stamp the right starting status field (`x_fc_adp_application_status` for ADP, `x_fc_mod_status` for MOD, etc. — mirroring `assessment.py:600-622`) and the right links. **Reuse the existing MOD/ADP/ODSP pipelines unchanged.**
### 3.7 Emails
- Reuse `fusion.email.builder.mixin` and the existing per-funding stage emails (they're keyed off SO type + status, so per-SO they keep working).
- **Move the completion send to per-SO** inside the new builder (not per-assessment), and **dedupe recipients**, so a 3-item visit doesn't emit 3–6 completion emails.
- **Fix the existing duplicate** (authorizer gets two completion emails on backend ADP completion) as part of this.
- Make `enable_email_notifications` gating consistent across the sends the visit touches.
### 3.8 Reused vs net-new
- **Reused, largely untouched:** the 7 accessibility measurement forms + their JS/Python calc; the ADP Express form + Page-11 signature; the MOD/ADP/ODSP pipelines, views, wizards, and stage emails; the email branding mixin.
- **Net-new:** the `fusion.assessment.visit` model + workspace UI; per-item funding selector on the accessibility forms; the group-aware SO builder + link-cardinality change; ADP multi-device + combination validation; scooter type + fields; power-mobility home-access rule + cross-sell flag; completion-email consolidation.
---
## 4. Resolved decisions
1.**MOD funding cap — documented rule, light-touch in v1.** March of Dimes covers **up to $15,000 per person, lifetime**, income-gated: if the client's income is **under** that year's threshold (the threshold changes annually), MOD funds the full $15k; if **over**, MOD may **deny or partially approve**. **v1:** surface this cap as a reminder on MOD items and capture an *"income under MOD threshold? (yes / no / unknown)"* flag so the rep can judge — **do not** auto-compute lifetime used-vs-remaining across the client's prior MOD orders (the SO's existing `x_fc_mod_*` approved/payment fields already record per-order amounts). **Future:** yearly-threshold config + automatic lifetime-remaining tracking + a hard warning.
2.**No auto pricing / products in v1.** The visit creates a **draft** SO per funding workflow and appends each assessment's specs to that SO's chatter (today's pattern); **the sales rep builds the quotation lines manually.** One SO can hold many items. No per-assessment-type product mapping. (Auto-pricing is a future expansion.)
3.**Patient-lift funding is chosen per case** via the funding picker — March of Dimes, **ODSP**, or **Hardship** (e.g. Toronto residents) all fund it; it is not fixed (see §3.4).
4.**Power-wheelchair form unchanged** — already well-optimized; the only addition is the **home-accessibility warning** (device usable **inside and outside** the home). The home-access rule applies to **scooter (new type, new section) and power wheelchair (warning only)**.
---
## 5. Phasing
- **Phase 1 — Funding correctness + visit backbone:** `fusion.assessment.visit`, link-cardinality change, **funding selector on the accessibility forms** (incl. Hardship; patient-lift routing), **MOD $15k-cap reminder + income-threshold flag** (informational), group-and-route to per-workflow **draft** SOs (specs to chatter, manual pricing) reusing existing pipelines, completion-email consolidation + duplicate fix. *(Delivers the MOD-routing fix and the multi-SO split.)*
- **Phase 2 — ADP expansion:** multi-device ADP order + combination validation, **scooter** type + fields, power-mobility **home-access hard rule** + accessibility cross-sell prompt.
- **Phase 3 — Seamless field UX:** the full add-as-you-go workspace, measurement-first deferral, location tags, "same as previous", OT on-site sign-off polish.
- **Later:** product-line auto-pricing, MOD funding-cap tracking, voice/quick entry.
---
## 6. Risks (from investigation)
- **Duplicate completion emails** already live on the ADP backend path — fix before fan-out (§3.7).
- **Scalar back-links + double-SO guards** assume one SO per assessment; grouping breaks them — must move to `visit_id` / One2many and make the guard visit-aware.
- **Inconsistent `enable_email_notifications`** — template sends ignore the kill-switch; don't route new traffic through templates without honoring it.
- **Label drift** `x_fc_funding_source` vs `x_fc_sale_type` (`insurance`="Private Insurance" vs "Insurance"; `direct_private`="Private Pay (Direct)" vs "Direct/Private") — keys match so routing works; align labels in any shared UI.
- **Unreachable funding types from accessibility:** `sale_type_map` (`accessibility_assessment.py:771`) covers 6 values; decide which funding types each assessment type may emit.
Nodes (20): Assign role-specific portal groups to a portal user based on contact checkboxes., Assign backend groups to an internal user based on contact checkboxes. A, Grant portal access to this partner, or update permissions for existing users., Create a role-specific welcome Knowledge article for the new portal user., Send a professional portal invitation email to the partner. Gen, Resend portal invitation email to an existing portal user., Open the list of assigned sale orders, Open the list of assessments for this partner (+12 more)
### Community 2 - "Community 2"
Cohesion: 0.07
Nodes (19): create(), FusionAssessment, Format assessment data as HTML table for chatter, Format wheelchair specifications for the sale order notes (legacy), Generate document records for signed pages, Send email notifications when assessment is completed, View related documents, View the created sale order (+11 more)
### Community 3 - "Community 3"
Cohesion: 0.08
Nodes (15): create(), FusionAccessibilityAssessment, Complete the assessment and create a Sale Order. 2026-04 portal audit f, Add a tag to the sale order based on assessment type, Copy assessment photos to sale order chatter, Send email notification to office about assessment completion, Schedule a follow-up activity for the sales rep, Find or create a partner for the client (+7 more)
### Community 4 - "Community 4"
Cohesion: 0.08
Nodes (20): Complete express assessment and create draft sale order (no signatures required), CustomerPortal, Ensure all module views are active after install/update. Odoo silently deac, _reactivate_views(), AssessmentPortal, portal_assessment_express_edit(), portal_assessment_express_new(), portal_assessment_express_save() (+12 more)
### Community 5 - "Community 5"
Cohesion: 0.09
Nodes (14): authorizer_cases_search(), sales_rep_cases_search(), get_authorizer_portal_cases(), get_sales_rep_portal_cases(), Open composer to send message to authorizer only, Send email when an authorizer is assigned to the order, View portal documents, Get data for portal display, excluding sensitive information (+6 more)
### Community 6 - "Community 6"
Cohesion: 0.12
Nodes (14): preview_pdf(), _draw_field(), fill_template(), PDFTemplateFiller, Generic PDF template filler. Works with any template, any number of pages., create(), FusionPdfTemplate, FusionPdfTemplateField (+6 more)
### Community 7 - "Community 7"
Cohesion: 0.11
Nodes (14): accessibility_assessment_save(), AuthorizerPortal, Portal controller for Authorizers (OTs/Therapists), Parse straight stair lift specific fields, Parse curved stair lift specific fields, Parse VPL specific fields, Parse ceiling lift specific fields, Parse ramp specific fields (+6 more)
Nodes (1): Transcribe voice recording using OpenAI Whisper, translate to English.
### Community 96 - "Community 96"
Cohesion: 1.0
Nodes (1): Use GPT to clean up and format raw notes text.
### Community 97 - "Community 97"
Cohesion: 1.0
Nodes (1): Format transcription with GPT and complete the task.
### Community 98 - "Community 98"
Cohesion: 1.0
Nodes (1): Next day preparation view.
### Community 99 - "Community 99"
Cohesion: 1.0
Nodes (1): View schedule for a specific date.
### Community 100 - "Community 100"
Cohesion: 1.0
Nodes (1): Admin map view showing latest technician locations using Google Maps.
### Community 101 - "Community 101"
Cohesion: 1.0
Nodes (1): Log the technician's current GPS location.
### Community 102 - "Community 102"
Cohesion: 1.0
Nodes (1): Check if the current technician is clocked in. Returns {clocked_in: boo
### Community 103 - "Community 103"
Cohesion: 1.0
Nodes (1): Save the technician's personal start location.
### Community 104 - "Community 104"
Cohesion: 1.0
Nodes (1): Register a push notification subscription.
### Community 105 - "Community 105"
Cohesion: 1.0
Nodes (1): Legacy: List of deliveries for the technician (redirects to tasks).
### Community 106 - "Community 106"
Cohesion: 1.0
Nodes (1): View a specific delivery for technician (legacy, still works).
### Community 107 - "Community 107"
Cohesion: 1.0
Nodes (1): POD signature capture page - accessible by technicians and sales reps
### Community 108 - "Community 108"
Cohesion: 1.0
Nodes (1): Save POD signature via AJAX
### Community 109 - "Community 109"
Cohesion: 1.0
Nodes (1): Task-level POD signature capture page (works for all tasks including shadow).
### Community 110 - "Community 110"
Cohesion: 1.0
Nodes (1): Save POD signature directly on a task.
### Community 111 - "Community 111"
Cohesion: 1.0
Nodes (1): Show the accessibility assessment type selector
### Community 112 - "Community 112"
Cohesion: 1.0
Nodes (1): List all accessibility assessments for the current user (sales rep or authorizer
### Community 113 - "Community 113"
Cohesion: 1.0
Nodes (1): Straight stair lift assessment form
### Community 114 - "Community 114"
Cohesion: 1.0
Nodes (1): Curved stair lift assessment form
### Community 115 - "Community 115"
Cohesion: 1.0
Nodes (1): Vertical Platform Lift assessment form
### Community 116 - "Community 116"
Cohesion: 1.0
Nodes (1): Ceiling Lift assessment form
### Community 117 - "Community 117"
Cohesion: 1.0
Nodes (1): Custom Ramp assessment form
### Community 118 - "Community 118"
Cohesion: 1.0
Nodes (1): Bathroom Modification assessment form
### Community 119 - "Community 119"
Cohesion: 1.0
Nodes (1): Tub Cutout assessment form
### Community 120 - "Community 120"
Cohesion: 1.0
Nodes (1): Save an accessibility assessment and optionally create a Sale Order
### Community 121 - "Community 121"
Cohesion: 1.0
Nodes (1): Render the rental pickup inspection form for the technician.
### Community 122 - "Community 122"
Cohesion: 1.0
Nodes (1): Save the rental inspection results.
## Knowledge Gaps
- **177 isolated node(s):** `Ensure all module views are active after install/update. Odoo silently deac`, `Generic PDF template filler. Works with any template, any number of pages.`, `Fill a PDF template by overlaying text/checkmarks/signatures at configured posit`, `Draw a single field onto the reportlab canvas. Args: c: rep`, `Override create to generate reference number` (+172 more)
These have ≤1 connection - possible missing edges or undocumented components.
- **Thin community `Community 19`** (1 nodes): `__init__.py`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 20`** (1 nodes): `__init__.py`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 21`** (1 nodes): `__init__.py`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 22`** (1 nodes): `__manifest__.py`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 23`** (1 nodes): `Fill a PDF template by overlaying text/checkmarks/signatures at configured posit`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 24`** (1 nodes): `Draw a single field onto the reportlab canvas. Args: c: rep`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 25`** (1 nodes): `Override create to generate reference number`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 26`** (1 nodes): `Get authorizer from x_fc_authorizer_id field`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 27`** (1 nodes): `Get cases for authorizer portal with optional search`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 28`** (1 nodes): `Get cases for sales rep portal with optional search`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 29`** (1 nodes): `Override create to handle revision numbering`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 30`** (1 nodes): `Get documents for a sale order, optionally filtered by type`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 31`** (1 nodes): `Get all revisions of a specific document type`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 32`** (1 nodes): `Override create to set author from current user if not provided`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 33`** (1 nodes): `Kanban group expansion — always show all 6 workflow states.`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 34`** (1 nodes): `Straight stair lift: (steps × nose_to_nose) + 13" top landing`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 35`** (1 nodes): `Use manual override if provided, otherwise use calculated`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 36`** (1 nodes): `Curved stair lift calculation: - 12" per step - 16" per curve`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 37`** (1 nodes): `Use manual override if provided, otherwise use calculated`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 38`** (1 nodes): `Ontario Building Code: 12 inches length per 1 inch height (1:12 ratio)`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 39`** (1 nodes): `Landing required every 30 feet (360 inches)`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 40`** (1 nodes): `Total length including landings (5 feet = 60 inches each)`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 41`** (1 nodes): `Compute portal access status based on user account and login history.`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 42`** (1 nodes): `Count sale orders where this partner is the authorizer`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 43`** (1 nodes): `Count assessments where this partner is involved`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 44`** (1 nodes): `Count sale orders assigned to this partner as delivery technician`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 45`** (1 nodes): `assessment_form.js`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 46`** (1 nodes): `technician_sw.js`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 47`** (1 nodes): `loaner_portal.js`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 48`** (1 nodes): `signature_pad.js`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 49`** (1 nodes): `portal_search.js`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 50`** (1 nodes): `Display the Page 11 signing form.`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 51`** (1 nodes): `Process the submitted Page 11 signature.`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 52`** (1 nodes): `Download the signed Page 11 PDF.`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 53`** (1 nodes): `Start a new assessment`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 54`** (1 nodes): `View/edit an assessment`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 55`** (1 nodes): `Save assessment data (create or update)`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 56`** (1 nodes): `Signature capture page`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 57`** (1 nodes): `Save a signature (AJAX)`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 58`** (1 nodes): `Complete the assessment`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 59`** (1 nodes): `Start a new express assessment (Page 1 - Equipment Selection)`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 60`** (1 nodes): `Continue/edit an express assessment`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 61`** (1 nodes): `Save express assessment data (create or update)`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 62`** (1 nodes): `Public page for booking an accessibility assessment.`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 63`** (1 nodes): `Process assessment booking form submission.`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 64`** (1 nodes): `Render the visual field editor for a PDF template.`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 65`** (1 nodes): `Return all fields for a template.`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 66`** (1 nodes): `Update a field's position or properties.`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 67`** (1 nodes): `Create a new field on a template.`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 68`** (1 nodes): `Delete a field from a template.`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 69`** (1 nodes): `Return the preview image URL for a specific page.`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 70`** (1 nodes): `Upload a preview image for a template page directly from the editor.`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 71`** (1 nodes): `Generate a preview filled PDF with sample data.`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 72`** (1 nodes): `Auto-save browser-detected timezone to the user profile if not already set.`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 73`** (1 nodes): `Override home to add ADP posting info for Fusion users`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 120`** (1 nodes): `Save an accessibility assessment and optionally create a Sale Order`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 121`** (1 nodes): `Render the rental pickup inspection form for the technician.`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
- **Thin community `Community 122`** (1 nodes): `Save the rental inspection results.`
Too small to be a meaningful cluster - may be noise or needs more connections extracted.
## Suggested Questions
_Questions this graph is uniquely positioned to answer:_
- **Why does `create()` connect `Community 3` to `Community 0`, `Community 1`, `Community 4`, `Community 7`, `Community 11`?**
_High betweenness centrality (0.080) - this node is a cross-community bridge._
- **Why does `FusionAssessment` connect `Community 2` to `Community 4`?**
_High betweenness centrality (0.059) - this node is a cross-community bridge._
- **Why does `AuthorizerPortal` connect `Community 7` to `Community 0`, `Community 4`?**
_High betweenness centrality (0.047) - this node is a cross-community bridge._
- **Are the 17 inferred relationships involving `create()` (e.g. with `._generate_tutorial_articles()` and `.action_grant_portal_access()`) actually correct?**
_`create()` has 17 INFERRED edges - model-reasoned connections that need verification._
- **Are the 2 inferred relationships involving `accessibility_assessment_save()` (e.g. with `create()` and `.action_complete()`) actually correct?**
_`accessibility_assessment_save()` has 2 INFERRED edges - model-reasoned connections that need verification._
- **What connects `Ensure all module views are active after install/update. Odoo silently deac`, `Generic PDF template filler. Works with any template, any number of pages.`, `Fill a PDF template by overlaying text/checkmarks/signatures at configured posit` to the rest of the system?**
_177 weakly-connected nodes found - possible documentation gaps or missing edges._
- **Should `Community 0` be split into smaller, more focused modules?**
_Cohesion score 0.05 - nodes in this community are weakly interconnected._
{"nodes":[{"id":"users_gurpreet_github_odoo_modules_fusion_authorizer_portal_models_authorizer_comment_py","label":"authorizer_comment.py","file_type":"code","source_file":"/Users/gurpreet/Github/Odoo-Modules/fusion_authorizer_portal/models/authorizer_comment.py","source_location":"L1"},{"id":"authorizer_comment_authorizercomment","label":"AuthorizerComment","file_type":"code","source_file":"/Users/gurpreet/Github/Odoo-Modules/fusion_authorizer_portal/models/authorizer_comment.py","source_location":"L9"},{"id":"authorizer_comment_compute_display_name","label":"_compute_display_name()","file_type":"code","source_file":"/Users/gurpreet/Github/Odoo-Modules/fusion_authorizer_portal/models/authorizer_comment.py","source_location":"L70"},{"id":"authorizer_comment_create","label":"create()","file_type":"code","source_file":"/Users/gurpreet/Github/Odoo-Modules/fusion_authorizer_portal/models/authorizer_comment.py","source_location":"L78"},{"id":"authorizer_comment_rationale_79","label":"Override create to set author from current user if not provided","file_type":"rationale","source_file":"/Users/gurpreet/Github/Odoo-Modules/fusion_authorizer_portal/models/authorizer_comment.py","source_location":"L79"}],"edges":[{"source":"users_gurpreet_github_odoo_modules_fusion_authorizer_portal_models_authorizer_comment_py","target":"odoo","relation":"imports_from","confidence":"EXTRACTED","source_file":"/Users/gurpreet/Github/Odoo-Modules/fusion_authorizer_portal/models/authorizer_comment.py","source_location":"L3","weight":1.0},{"source":"users_gurpreet_github_odoo_modules_fusion_authorizer_portal_models_authorizer_comment_py","target":"logging","relation":"imports","confidence":"EXTRACTED","source_file":"/Users/gurpreet/Github/Odoo-Modules/fusion_authorizer_portal/models/authorizer_comment.py","source_location":"L4","weight":1.0},{"source":"users_gurpreet_github_odoo_modules_fusion_authorizer_portal_models_authorizer_comment_py","target":"authorizer_comment_authorizercomment","relation":"contains","confidence":"EXTRACTED","source_file":"/Users/gurpreet/Github/Odoo-Modules/fusion_authorizer_portal/models/authorizer_comment.py","source_location":"L9","weight":1.0},{"source":"users_gurpreet_github_odoo_modules_fusion_authorizer_portal_models_authorizer_comment_py","target":"authorizer_comment_compute_display_name","relation":"contains","confidence":"EXTRACTED","source_file":"/Users/gurpreet/Github/Odoo-Modules/fusion_authorizer_portal/models/authorizer_comment.py","source_location":"L70","weight":1.0},{"source":"users_gurpreet_github_odoo_modules_fusion_authorizer_portal_models_authorizer_comment_py","target":"authorizer_comment_create","relation":"contains","confidence":"EXTRACTED","source_file":"/Users/gurpreet/Github/Odoo-Modules/fusion_authorizer_portal/models/authorizer_comment.py","source_location":"L78","weight":1.0},{"source":"authorizer_comment_rationale_79","target":"authorizer_comment_authorizercomment_create","relation":"rationale_for","confidence":"EXTRACTED","source_file":"/Users/gurpreet/Github/Odoo-Modules/fusion_authorizer_portal/models/authorizer_comment.py","source_location":"L79","weight":1.0}],"raw_calls":[{"caller_nid":"authorizer_comment_compute_display_name","callee":"strftime","source_file":"/Users/gurpreet/Github/Odoo-Modules/fusion_authorizer_portal/models/authorizer_comment.py","source_location":"L73"},{"caller_nid":"authorizer_comment_compute_display_name","callee":"_","source_file":"/Users/gurpreet/Github/Odoo-Modules/fusion_authorizer_portal/models/authorizer_comment.py","source_location":"L75"},{"caller_nid":"authorizer_comment_create","callee":"get","source_file":"/Users/gurpreet/Github/Odoo-Modules/fusion_authorizer_portal/models/authorizer_comment.py","source_location":"L81"},{"caller_nid":"authorizer_comment_create","callee":"get","source_file":"/Users/gurpreet/Github/Odoo-Modules/fusion_authorizer_portal/models/authorizer_comment.py","source_location":"L83"},{"caller_nid":"authorizer_comment_create","callee":"super","source_file":"/Users/gurpreet/Github/Odoo-Modules/fusion_authorizer_portal/models/authorizer_comment.py","source_location":"L85"}]}
{"nodes":[{"id":"users_gurpreet_github_odoo_modules_fusion_authorizer_portal_init_py","label":"__init__.py","file_type":"code","source_file":"/Users/gurpreet/Github/Odoo-Modules/fusion_authorizer_portal/__init__.py","source_location":"L1"},{"id":"init_reactivate_views","label":"_reactivate_views()","file_type":"code","source_file":"/Users/gurpreet/Github/Odoo-Modules/fusion_authorizer_portal/__init__.py","source_location":"L7"},{"id":"init_rationale_8","label":"Ensure all module views are active after install/update. Odoo silently deac","file_type":"rationale","source_file":"/Users/gurpreet/Github/Odoo-Modules/fusion_authorizer_portal/__init__.py","source_location":"L8"}],"edges":[{"source":"users_gurpreet_github_odoo_modules_fusion_authorizer_portal_init_py","target":"users_gurpreet_github_odoo_modules_fusion_authorizer_portal_init_py","relation":"imports_from","confidence":"EXTRACTED","source_file":"/Users/gurpreet/Github/Odoo-Modules/fusion_authorizer_portal/__init__.py","source_location":"L3","weight":1.0},{"source":"users_gurpreet_github_odoo_modules_fusion_authorizer_portal_init_py","target":"users_gurpreet_github_odoo_modules_fusion_authorizer_portal_init_py","relation":"imports_from","confidence":"EXTRACTED","source_file":"/Users/gurpreet/Github/Odoo-Modules/fusion_authorizer_portal/__init__.py","source_location":"L4","weight":1.0},{"source":"users_gurpreet_github_odoo_modules_fusion_authorizer_portal_init_py","target":"init_reactivate_views","relation":"contains","confidence":"EXTRACTED","source_file":"/Users/gurpreet/Github/Odoo-Modules/fusion_authorizer_portal/__init__.py","source_location":"L7","weight":1.0},{"source":"init_rationale_8","target":"init_reactivate_views","relation":"rationale_for","confidence":"EXTRACTED","source_file":"/Users/gurpreet/Github/Odoo-Modules/fusion_authorizer_portal/__init__.py","source_location":"L8","weight":1.0}],"raw_calls":[{"caller_nid":"init_reactivate_views","callee":"search","source_file":"/Users/gurpreet/Github/Odoo-Modules/fusion_authorizer_portal/__init__.py","source_location":"L15"},{"caller_nid":"init_reactivate_views","callee":"sudo","source_file":"/Users/gurpreet/Github/Odoo-Modules/fusion_authorizer_portal/__init__.py","source_location":"L15"},{"caller_nid":"init_reactivate_views","callee":"write","source_file":"/Users/gurpreet/Github/Odoo-Modules/fusion_authorizer_portal/__init__.py","source_location":"L20"},{"caller_nid":"init_reactivate_views","callee":"execute","source_file":"/Users/gurpreet/Github/Odoo-Modules/fusion_authorizer_portal/__init__.py","source_location":"L21"},{"caller_nid":"init_reactivate_views","callee":"fetchall","source_file":"/Users/gurpreet/Github/Odoo-Modules/fusion_authorizer_portal/__init__.py","source_location":"L26"},{"caller_nid":"init_reactivate_views","callee":"warning","source_file":"/Users/gurpreet/Github/Odoo-Modules/fusion_authorizer_portal/__init__.py","source_location":"L28"},{"caller_nid":"init_reactivate_views","callee":"getLogger","source_file":"/Users/gurpreet/Github/Odoo-Modules/fusion_authorizer_portal/__init__.py","source_location":"L28"},{"caller_nid":"init_reactivate_views","callee":"len","source_file":"/Users/gurpreet/Github/Odoo-Modules/fusion_authorizer_portal/__init__.py","source_location":"L29"}]}
File diff suppressed because it is too large
Load Diff
Some files were not shown because too many files have changed in this diff
Show More
Reference in New Issue
Block a user
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.