feat(harness): signal-impact ablation tool + deterministic-signal conventions (5.3.0)#499
Merged
Conversation
…ventions Adds scripts/signal-impact.mjs (npm run benchmark:signal-impact): joins a labeled manifest to its local text, runs analyzeText() per row, and recomputes the hot verdict with each signal ablated to report each signal's marginal catch/FP (attributable TP/FP) + Δrecall/Δfpr/ΔF1. Deterministic and gitignore-safe. Replaces the ad-hoc eval-kernel measurement used for the 5.2.0 KO signal; unit-tested with synthetic inputs. Adds docs/HARNESS.md (index of every measurement/calibration/gate tool) and a "Adding a Deterministic Detection Signal" convention to CONTRIBUTING.md + CONTRIBUTING_KR.md (diagnose -> FP-safe discriminator -> first-class impl -> mirror every surface -> measure with the harness -> version). Links from README*, tests/quality/README.md.
Minor bump for the contributor-facing signal-impact harness + deterministic detection-signal conventions. No CLI/schema/pattern/detection-behavior change; all four languages byte-identical. Syncs version surfaces + CHANGELOG.
Contributor
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Defines and consolidates patina's measurement harness and signal-authoring conventions after a burst of additions. Bumps to 5.3.0 (minor — contributor-facing tooling + workflow; no CLI/schema/detection-behavior change, all four languages byte-identical).
What's added
Signal-impact harness —
scripts/signal-impact.mjs(npm run benchmark:signal-impact)expected_hot) to its local text, runsanalyzeText()once per row, then recomputes the document hot verdict with each signal ablated to report each signal's marginal contribution: attributable TP/FP (rows it alone keeps hot) + Δrecall/Δfpr/ΔF1, plus catch-by-model-family.recomputeHot()mirrors theanalyzeTextOR rule; a unit test pins the signal list.--ablate ko_ending_monotony→ attributable TP 15 / FP 3, Δrecall +11.6, ΔF1 +0.072.Harness map —
docs/HARNESS.mdREADME*andtests/quality/README.md.Convention —
CONTRIBUTING.md+CONTRIBUTING_KR.md"Adding a Deterministic Detection Signal"Verification
npm test— 791 pass / 0 fail (4 new signal-impact unit tests)npm run lint— syntax OK (159 files), cspell 0 issuesnpm run release:check— OK for 5.3.0npm run check:no-private-assets— OK (311 packed)npm run benchmark— 49-fixture suite still 100%