hamid
b8934f531d
Three domain services (mirroring the patients/nurse template): services/geography (cached province→city→district lookups; Infinity staleTime + shared geographyKeys), services/addresses (address book CRUD + set-primary; single-primary invariant), and services/serviceAreas (coverage add/remove; areaExists dup-guard, districtId=null = whole city). Four tested composites in src/components/geography: CascadingRegionSelect (drives the cascade queries), AddressMapPicker (map-pin stand-in emitting real lat/lng), AddressForm, AddressCard. Screens: customer address book (/addresses, reached from the profile hub) and nurse coverage editor (/nurse/coverage, new sidebar tab, inline duplicate block + 409). Adds geo/address/coverage i18n namespaces (both locales), location/delete/coverage icons, ADDRESSES/NURSE_COVERAGE routes. Consumes the b4 geography-addresses contract; filed REQ-008 (accept the map pin on create/update) and REQ-009 (provinceId on CustomerAddressDto) for gaps. Gate: npm run check + npm run test:ci (129, +17) + npm run build all green. A 5-dimension adversarial review fixed 3 findings (map-marker RTL transform, page_size→pageSize pagination casing, coverage districts-scope dead-end on district-less cities). Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Phase reports — template
Every phase writes a report here when it finishes (operating-rules §7, Definition of Done). The report is the durable answer to "what did this phase do, what can I test now, and what's still fake?" — saved to a file, not left in chat.
File name: backend-phase-N-report.md or frontend-phase-N-bM-report.md.
# <Backend|Frontend> Phase N — <Title> — Report (<YYYY-MM-DD>)
## What was built
- Bullet list of concrete deliverables (entities/migrations/endpoints, or screens/services/components).
## What is now testable (and exactly how)
- Step-by-step for a human: e.g. "Swagger → POST /api/v1/auth/otp/request with {phone} → 200; the OTP is
logged to the console; POST /api/v1/auth/otp/verify with that code → returns access+refresh tokens."
- For the frontend: which screen, which route, what to click, expected result, how mock data shows up.
## What is mocked / waiting on a real service
- Each seam touched: interface + file, what's faked, and a link to its entry in `mocks-registry.md`.
## Contracts
- Produced (backend): which `contracts/domains/*.md` + openapi snapshot.
- Consumed (frontend): which contract/version; any request filed in `frontend/requests/for-backend.md`.
## Docs updated
- Which CLAUDE.md / product doc / conventions were updated and why.
## Follow-ups for later phases
- Anything intentionally deferred, with the phase that should pick it up.