Repository navigation
fix: Bugbot reconciliation retains an orphan unknown finding alongside duplicate semantic reports (#655) - #658
Conversation
…ongside-duplicate-semantic-reports: preserve durable finding identity during reconciliation
|
To preview the documentation for this pull request, visit the following URL:
|
There was a problem hiding this comment.
Current status: All findings originating in this review are resolved.
Last reconciled onc02a600. See aggregate Bugbot status.
🤖 Bugbot review snapshot
Bugbot reported 2 potential problems when commit cb0ed01 was analyzed. This snapshot is historical; use the status block above for current state. 2 findings are linked to changed code.
Findings
- medium: Rekeying can give distinct findings the same ID —
src/domain/bugbot/finding.ts:125 - low: Unknown finding IDs can inject Markdown links —
src/application/policies/bugbot_review_presentation_policy.ts:239
Bugbot: review complete
|
Copilot approval · Human review neededStatus: A protected workflow, policy, or trust-boundary file changed. Technical evidenceRevision: |
Codecov Report❌ Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## develop #658 +/- ##
===========================================
+ Coverage 90.94% 94.13% +3.19%
===========================================
Files 135 972 +837
Lines 6251 31514 +25263
Branches 1338 9247 +7909
===========================================
+ Hits 5685 29666 +23981
- Misses 172 715 +543
- Partials 394 1133 +739
🚀 New features to boost your workflow:
|
…ongside-duplicate-semantic-reports: preserve colliding reports and render unknown IDs as code
Refreshing a fingerprint-matched Bugbot finding could replace its durable ID with the model's new ID, leaving a second, unlinked
unknownrecord in the final projection. This change keeps the original ID and native thread while refreshing the current evidence.What changed
Validation
graphify update .completed.Review notes
The replay demonstrates the implementation defect without asserting a historical consumer API outage cause. Reference predictions validate the scorer; they do not establish real-model accuracy improvement. Unpublished candidates remain transient attempt evidence, and their later disappearance does not prove resolution; recovery documentation states that boundary explicitly. No review gate, permission or confidence threshold is weakened.
Closes #655