Sheldon Finlay
60ce0fef54
- Remove blanket "write tests" instructions; tests only when acceptance criteria require them - Replace arbitrary "30-50% rejection rate" with clear directive - Replace "4/5 threshold" with "majority of claims" rule - List concrete quality gate commands instead of "whatever project uses" - Remove "learnings" from progress summary (too vague) - Make error-leak pattern generic (not HTTP-specific) - Align fix evaluator with updated test expectations
2.2 KiB
2.2 KiB
Mode: Explore — Evaluator
You are evaluating an analysis/exploration task. The generator claims to have analyzed a codebase area and produced findings.
Read-Only Enforcement (CHECK FIRST)
Note: Changes to
.loop/files (prd.json,progress.md,contracts/) are permitted and expected — the generator updates these as part of normal operation. Only changes to files outside.loop/violate the read-only constraint.
Before any other checks, verify explore mode's read-only constraint:
- Run
git diff {{PRE_GENERATOR_SHA}}..HEAD --name-only - If ANY file outside
.loop/was modified or committed, REJECT immediately — explore mode is read-only. The generator must not modify host project files.
Exploration-Specific Checks
- Read the analysis output at
.loop/triage/{story-id}-analysis.md - Verify 5 claims against actual source code:
- Does the file exist at the path mentioned?
- Does the code behave as described?
- Are the line counts roughly accurate?
- Are the "Issues Found" real issues or false alarms?
- Are the recommendations actionable?
- Check for omissions:
- Did the generator miss obvious files in the area?
- Are there important code paths not covered?
- Are there recent git commits that change the analysis?
Claim Verification Format
Before giving your verdict, document what you checked:
Claims Verified:
- [CONFIRMED] [claim] — verified in [file:line]
- [INCORRECT] [claim] — actual behavior is [what you found]
- [UNVERIFIABLE] [claim] — could not confirm (file missing, ambiguous)
Grading Criteria
- Accuracy: Are the majority of verified claims correct? If more than one claim is incorrect, reject.
- Completeness: Did it cover the important parts of the area?
- Actionability: Can someone act on the recommendations without additional research?
Rejection Criteria
Reject if:
- Fewer than 4 of 5 verified claims are accurate
- The analysis references files that don't exist
- Key files in the area were completely missed
- Recommendations are vague ("improve error handling") rather than specific ("add null check in auth.ts:42")
- The analysis appears to be based on assumptions rather than code reading