Issues

• Field test: harness_health.json has no run identity — stale and truncated sweeps report as passing

  bugfield-test
  deghosal-2026/CauterRule#818 · Filter by author deghosal-2026 opened on Sep 17, 2026

• Bug: Done: line divides the processed count by the non-quarantined target — numerator includes quarantined records

  bug
  deghosal-2026/CauterRule#817 · Filter by author deghosal-2026 opened on Sep 17, 2026

• Field test: precision uses the source-domain pool, hiding cross-domain over-fire (recall and precision need different pools)

  enhancementfield-test
  deghosal-2026/CauterRule#816 · Filter by author deghosal-2026 opened on Sep 17, 2026

• Bug: same-domain git over-fire — push-specific rule still marks git status/git pull successes as broken

  bug
  deghosal-2026/CauterRule#815 · Filter by author deghosal-2026 opened on Sep 17, 2026

• Feature: counterfactual hard negatives (high token overlap, different failure class) + structural artifact holdout

  enhancementfield-test
  deghosal-2026/CauterRule#814 · Filter by author deghosal-2026 opened on Sep 15, 2026

• Field test: score-preserving adversarial candidate generation (keep the score, break the meaning)

  enhancementfield-test
  deghosal-2026/CauterRule#812 · Filter by author deghosal-2026 opened on Sep 15, 2026

• Field test: hold out formats/tool vocabularies, not just examples, in replay evaluation

  enhancementfield-test
  deghosal-2026/CauterRule#813 · Filter by author deghosal-2026 opened on Sep 15, 2026

• Field test: generalize degenerate-trigger detection — negative-corpus invariant beyond artifact blacklist

  enhancementfield-test
  deghosal-2026/CauterRule#811 · Filter by author deghosal-2026 opened on Sep 15, 2026

• Bug: CORPUS_THRESHOLDS in run-field-test.py is dead code and misleads threshold reasoning

  bug
  deghosal-2026/CauterRule#810 · Filter by author deghosal-2026 opened on Sep 15, 2026

• Field test: per-trajectory timeout telemetry (counts, rates, per-corpus distribution)

  enhancementfield-test
  deghosal-2026/CauterRule#809 · Filter by author deghosal-2026 opened on Sep 15, 2026

• Feature: add a regression test pinning the per-trajectory watchdog (and correct the article's as_completed snippet)

  enhancement
  deghosal-2026/CauterRule#808 · Filter by author deghosal-2026 opened on Sep 15, 2026

• Bug: _is_transient() retries timeouts, contradicting documented #713 non-retryable-timeout behavior

  bug
  deghosal-2026/CauterRule#807 · Filter by author deghosal-2026 opened on Sep 15, 2026

• Bug: HTTP 413 / TPM-metered token-quota rejections treated as terminal, not transient

  bug
  deghosal-2026/CauterRule#806 · Filter by author deghosal-2026 opened on Sep 15, 2026

• OpenSSF Scorecard remediation — branch protection, code review, fuzzing, contributed posture (<7/10)

  enhancement
  deghosal-2026/CauterRule#713 · Filter by author deghosal-2026 opened on Sep 12, 2026

• [v0.4.0] Human-vs-replay agreement rate (#493) — sample candidates per verdict bucket

  enhancement
  deghosal-2026/CauterRule#684 · Filter by author deghosal-2026 opened on Sep 11, 2026 · v0.4.0-M1: Deep Integrations

• [v0.4.0] Cross-session repeat-failure reduction (#663/#496) — 5-session protocol measurement

  enhancement
  deghosal-2026/CauterRule#683 · Filter by author deghosal-2026 opened on Sep 11, 2026 · v0.4.0-M1: Deep Integrations

• [v0.3.0·M7→v0.4.0] Cost measurement (#653) — extract token usage from cloud runs, publish $/1k per model

  enhancement
  deghosal-2026/CauterRule#682 · Filter by author deghosal-2026 opened on Sep 11, 2026 · v0.4.0-M1: Deep Integrations

• [v0.3.0·M7→v0.4.0] Semantic matching — embedding-based trigger↔reference similarity to bridge paraphrase gap

  enhancement
  deghosal-2026/CauterRule#680 · Filter by author deghosal-2026 opened on Sep 11, 2026 · v0.4.0-M1: Deep Integrations

• [v0.4.0·M6] Article ideas and announcements — dev.to articles, blog posts, social content

  enhancementv0.4.0
  deghosal-2026/CauterRule#672 · Filter by author deghosal-2026 opened on Sep 9, 2026 · v0.4.0-M6: Release Readiness

• [v0.4.0·M6] Merge all changes to main — final PR, code review, merge, verify CI

  enhancementv0.4.0
  deghosal-2026/CauterRule#669 · Filter by author deghosal-2026 opened on Sep 9, 2026 · v0.4.0-M6: Release Readiness

• [v0.4.0·M6] Update README badges — fresh v0.4.0 badges

  enhancementv0.4.0
  deghosal-2026/CauterRule#666 · Filter by author deghosal-2026 opened on Sep 9, 2026 · v0.4.0-M6: Release Readiness

• [v0.4.0·M6] Post-release checklist — verify PyPI, GitHub, Docker, milestones, announcements

  enhancementv0.4.0
  deghosal-2026/CauterRule#662 · Filter by author deghosal-2026 opened on Sep 9, 2026 · v0.4.0-M6: Release Readiness

• [v0.4.0·M5] M5 exit gate — code review, lint strict, coverage, docs updated, all issues resolved

  enhancementfield-testv0.4.0
  deghosal-2026/CauterRule#664 · Filter by author deghosal-2026 opened on Sep 9, 2026 · v0.4.0-M5: Field Test

• [v0.4.0·M6] Pre-release aggregate gate — verify all gates pass before tagging

  enhancementv0.4.0
  deghosal-2026/CauterRule#660 · Filter by author deghosal-2026 opened on Sep 9, 2026 · v0.4.0-M6: Release Readiness

• [v0.4.0·M5] Document known issues — per-feature severity, workaround, assignee

  enhancementfield-testv0.4.0
  deghosal-2026/CauterRule#659 · Filter by author deghosal-2026 opened on Sep 9, 2026 · v0.4.0-M5: Field Test