Repository navigation
Guide agents through implementation review after REV denials - #3
Merged
Merged
Conversation
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
REV denials currently expose scores and evidence relationships without enough direction on what to investigate next. An agent can mistake passing tests or a raw satisfied verdict for proof that only the evaluator needs fixing.
Add deterministic guidance to JSON reports and job summaries, plus a log pointer. Unresolved criteria identify missing accepted implementation or behavioral support and direct agents to trace production paths, investigate relevant bypasses, and reproduce confirmed defects with failing-before/passing-after regressions. Blocked evaluations are distinguished from implementation judgments. The agent contract documents review before escalation, including the form-validation example.
Guidance is advisory: evaluator prompts, thresholds, and approval decisions are unchanged. It discourages blind retries, weakened criteria, and unjustified guards.
Validation: npm run check and npm test passed (42 tests), including CLI report integration, evidence gaps, low-confidence satisfied choices, contradictions, infrastructure errors, approval, and escaping. Live model evaluation was not run because a TypeSafe API key is unavailable locally.