Build and Push Docker Image / build (push) Successful in 11m24sDetails
- Guard onboarding tRPC queries with session hydration check (fixes UNAUTHORIZED on first login)
- Defer expensive queries on awards page until UI elements are opened (dialog/tab)
- Fix perPage: 500 exceeding backend Zod max of 100 on awards eligibility query
- Add smooth open/close animation to project filters collapsible bar
- Fix seeded user status from ACTIVE to INVITED in seed-candidatures.ts
- Add router.refresh() cache invalidation across ~22 admin forms
- Fix geographic analytics query to use programId instead of round.programId
- Fix dashboard queries to scope by programId correctly
- Fix project.listPool and round queries for projects outside round context
- Add rounds page useEffect for state sync after mutations
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Build and Push Docker Image / build (push) Successful in 10m4sDetails
Special Awards:
- Add delete button with confirmation dialog to award detail page
- Add voting window dates (start/end) to award edit page
- Add manual project eligibility management (add/remove projects)
- Show eligibility method (Auto/Manual) in eligibility table
- Auto-set votingStartAt when opening voting if date is in future
Assignment Suggestions:
- Replace toggle with proper tabs UI (Algorithm vs AI Powered)
- Persist AI suggestions when navigating away (stored in database)
- Show suggestion counts on tab badges
- Independent refresh/start buttons per tab
Round Voting:
- Auto-update votingStartAt to now when activating round if date is in future
- Fixes issue where round was opened but voting dates were in future
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Build and Push Docker Image / build (push) Successful in 8m50sDetails
- Add AIUsageLog table migration for token tracking
- Fix GPT-5 temperature parameter (not supported, like o-series)
- Add usesNewTokenParam() and supportsTemperature() functions
- Add GPT-5+ category to model selection UI
- Update model sorting to show GPT-5+ first
GPT-5 and newer models use max_completion_tokens and don't support
custom temperature values, similar to reasoning models.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Matt