← overview
polyphony
9 signals · last run 11h ago · RTrentJones/Polyphony
| Signal | Kind | Latest model | Trend | Latest | Freshness | Commit |
|---|---|---|---|---|---|---|
| eval · cel · ci | quality | claude-opus-5-5 | ✓ 100% | 3d ago | b46d695 | |
| eval · dracula · ci | quality | gemini-3.5-flash-lite | ✗ 0% | 11h ago | 90363f2 | |
| eval · frankenstein · ci | quality | gemini-3.5-flash-lite | ✗ 0% | 12h ago | 90363f2 | |
| eval · wuthering · ci | quality | gemini-3.5-flash-lite | ✗ 0% | 11h ago | 90363f2 |
5 stale signals (no recent run; not counted in health)
| Signal | Kind | Latest model | Trend | Latest | Freshness | Commit |
|---|---|---|---|---|---|---|
| eval · ci-eval-backoff-cel | quality | gemini-2.5-flash | ✗ 0% | ◷ 14d ago | — | |
| eval · ci-eval-diag-cel | quality | gemini-2.5-flash | ✗ 0% | ◷ 14d ago | — | |
| eval · ci-judge-fix-cel | quality | gemini-2.5-flash | ✓ 100% | ◷ 14d ago | — | |
| eval · ci-post-reset-cel | quality | gemini-2.5-flash | ✗ 0% | ◷ 14d ago | — | |
| eval · ci-spec-merge-cel | quality | gemini-2.5-flash | ✗ 0% | ◷ 14d ago | — |
eval · frankenstein · ci (28 recent runs · all runs · compare models)
Pass rate per run
Metrics
Numeric results reported with each run, one chart per metric so each keeps its own scale.
attribution.accuracy0.267
attribution.chance0.333
attribution.score0.533
continuity.detection_recall1
continuity.false_positive_rate0
continuity.score1
extraction.f10.636
extraction.precision0.467
extraction.recall1
extraction.score0.636
ingestion.score0.643
outline.score0.880