A Switch at the top of the project side panel writes
useBalancedRanking onto Round.configJson via the existing round.update
mutation. The flip is shared across all viewers because the value
lives in the round's persisted config; hydration runs on every
roundData refetch so the UI converges quickly when another admin
flips it.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The admin dashboard fetches its side-sheet detail from project.getFullDetail
(not analytics.getProjectDetail as the audit assumed), and that procedure
had the same cross-round contamination bug. Add an optional roundId to
its input, filter the SUBMITTED-evaluations query when provided, and pass
roundId from the dashboard's useQuery so the Avg Score / Pass Rate /
Evaluators card now matches the per-juror list rendered below it.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Initial ranking order, advancement cutoff line, and per-row "advancing"
highlight now all use the juror-balanced (z-score corrected) average,
falling back to the raw avgGlobalScore when no balanced score exists.
The init effect waits for evalScores so the sort has the data it needs.
Admin drag-reorders still take precedence — saved reorders override the
default sort exactly as before.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Adds a shared juror-balancing utility (z-score normalization per juror,
rescaled back onto the raw 1-10 scale) and wires it into:
- Admin reports page: Top-10 project table now shows "Raw Avg" and
"Balanced" columns side by side, and the summary stats row shows a
balanced-average tile. Sort defaults to balanced so harsh and lenient
graders no longer skew the ranking.
- Ranking dashboard: each project row shows a green/amber balanced-score
chip next to the raw average when the two differ by ≥0.05, making it
obvious when juror calibration moved a project's effective ranking.
Also adds AI Juror Calibration Advisory — a mutation that takes
anonymized per-juror stats, calls OpenAI, and produces a plain-language
explanation of the cohort's grading patterns plus per-juror severity
(normal / notable / outlier) with a one-sentence narrative. The advisory
describes the statistical balance that already runs; it does not
introduce a new weighting layer. Rendered as a panel in the Juror
Consistency tab when a specific round is selected.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
- Fix project list returning empty for users with both SUPER_ADMIN and
JURY_MEMBER roles (jury filter now skips admins) in project, assignment,
and evaluation routers
- Add CountryDisplay component showing flag emoji + name everywhere
country is displayed (admin, observer, jury, mentor views — 17 files)
- Add countdown deadline banner on applicant dashboard for INTAKE,
SUBMISSION, and MENTORING rounds with live timer
- Remove quick action buttons from applicant dashboard
- Fix competition timeline sidebar: green dots/connectors only up to
current round, yellow dot for current round, red connector into
rejected round, grey after
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- processRoundClose now applies reordersJson drag-reorder overrides
when building the evaluation pass set (was ignoring admin reorders)
- Finalization tab groups proposed outcomes by category (Startup/Concept)
with per-group pass/reject/total counts
- Added category filter dropdown alongside the existing outcome filter
- Removed legacy "Advance Top N" button and dialog from ranking page
(replaced by the finalization workflow)
- Fix project edit status defaultValue showing empty placeholder
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Always display yes/no count format (e.g. "0/2 Yes") instead of
generic "2 jurors" when no advance votes exist
- Color coding: 2/2 Yes = green, 1/2 Yes = amber, 0/2 Yes = red
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
The display was sorted by compositeScore (which factors in pass rate and
z-score normalization) but the cutoff line and displayed values use
avgGlobalScore. This caused projects with high averages but low pass
rates (e.g. 1/2 Yes) to appear below the cutoff even when their average
exceeded the threshold.
Now sorts by avgGlobalScore (the visible metric) with compositeScore as
tiebreaker. Also adds a green left border to advancing projects for
clearer threshold highlighting.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Sort deduped ranking entries by compositeScore descending during
localOrder init — ensures correct display order for both formula
and old AI snapshots
- Fix threshold cutoff: scan forward to find first non-qualifying
project instead of backward scan that left non-qualifying projects
above the line
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Add scoreWeight and passRateWeight (0-10) to evaluation config for
configurable composite score formula. When ranking criteria text is
empty, triggerAutoRank uses pure formula ranking (no LLM calls).
When criteria text is present, AI-assisted ranking runs as before.
- Add FORMULA to RankingMode enum with migration
- Extract fetchCategoryProjects helper, add formulaRank service
- Update computeCompositeScore to accept configurable weights
- Add score/pass-rate weight sliders to ranking dashboard UI
- Mode-aware button labels (Calculator/formula vs Sparkles/AI)
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Create snapshot with status RUNNING before AI call starts
- Update to COMPLETED/FAILED when done
- Dashboard derives rankingInProgress from server snapshot status
- All admins see the spinner, not just the one who triggered it
- Poll snapshots every 3s so progress updates appear quickly
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Sort all ranked projects by compositeScore descending so highest-rated
projects always appear first (instead of relying on AI's inconsistent rank order)
- Deduplicate AI ranking response (AI sometimes returns same project multiple times)
- Deduplicate ranking entries and reorder IDs on dashboard load as defensive measure
- Show advancement cutoff line only once (precompute last advancing index)
- Override badge only shown when admin has actually drag-reordered (not on fresh rankings)
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Apply saved reordersJson when initializing the dashboard so any admin
sees the latest drag-reorder state, not just the original AI order.
The latest reorder event per category is used as the initial localOrder.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- AI ranking now includes ALL projects (never filters/excludes any)
- Updated system prompt: filter criteria inform priority, not exclusion
- Dynamic maxTokens scaling for large project pools (80 tokens/project)
- Fallback: projects AI omits are appended sorted by composite score
- Override badge uses snapshotOrder state (synced with localOrder in same
useEffect) instead of rankingMap.originalIndex to prevent stale-render
mismatch where all items incorrectly showed as overridden
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Header shows "Score >= X advance" instead of "Top N" in threshold mode
- Cutoff line placed after last project meeting threshold, not at fixed count
- Projects advancing determined by avg score vs threshold, not position
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Add criteriaWeights to EvaluationConfig for per-criterion weight assignment (0-10)
- Rewrite ai-ranking service: fetch eval form criteria, compute per-criterion averages,
z-score normalize juror scores to correct grading bias, send weighted criteria to AI
- Update AI prompts with criteria_definitions and per-project criteria_scores
- compositeScore uses weighted criteria when configured, falls back to globalScore
- Add collapsible ranking config section to dashboard (criteria text + weight sliders)
- Move rankingCriteria textarea from eval config tab to ranking dashboard
- Store criteriaWeights in ranking snapshot parsedRulesJson for audit
- Enhance projectScores CSV export with per-criterion averages, category, country
- Add Export CSV button to ranking dashboard header
- Add threshold-based advancement mode (decimal score threshold, e.g. 6.5)
alongside existing top-N mode in advance dialog
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Add adminEditEvaluation mutation and getJurorEvaluations query
- Create shared EvaluationEditSheet component with inline feedback editing
- Add Evaluations tab to member detail page (grouped by round)
- Make jury group member names clickable (link to member detail)
- Replace inline EvaluationDetailSheet on project page with shared component
- Fix project status transition validation (skip when status unchanged)
- Fix frontend to not send status when unchanged on project edit
- Ranking dashboard improvements and boolean decision converter fixes
- Backfill script updates for binary decisions
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Adds getProjectStates query to build a projectInfoMap lookup. Each row now
shows the project title, team name, and country. Also renames "Pass" label
to "Yes" to match the binary yes/no vote semantics.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Add pendingReorderCount ref + onMutate/onSettled to saveReorderMutation (DASH-07)
- Add advanceMutation (trpc.round.advanceProjects) with getProjectStates invalidation
- Add batchRejectMutation (trpc.roundEngine.batchTransition) using .length per MEMORY.md
- Add handleAdvance: advances top N per category, optionally batch-rejects the rest
- Add Advance Top N button in header (disabled when saveReorderMutation.isPending)
- Add Dialog with per-category N inputs, batch-reject checkbox, and count preview
- Import Dialog, Input, Label, Trophy from shadcn/lucide
