Processing Bot Reviews

Systematically process AI bot review findings.

Review Processing Steps

1. •Verify - Confirm the issue exists and is worth fixing
2. •Prioritize - Focus on real issues, skip false positives
3. •Fix - Apply fixes with tests
4. •Document - Update docs if needed

Workflow

When given review findings:

1. •Read each finding carefully
2. •Check if the issue exists in current code
3. •Determine if it's real or a false positive
4. •Fix real issues, note false positives
5. •Run lint/tests before pushing

Breaking Review Cycles

If bots keep flagging the same "fixed" issues:

1. •Pre-filter: "Review these findings and identify which are real vs false positives"
2. •After 2-3 cycles: Manually verify the PR diff
3. •Stop when:
   • •No new findings, OR
   • •All remaining are verified false positives

Good Finding Format

Good:

src/auth.ts:34-56 - Function doesn't handle null values.
Add null check before accessing user.email.

Bad:

The code could be better.

Output Format

When reporting on review processing:

## Review Summary

**Fixed:**
- src/auth.ts:34 - Added null check for user.email

**False Positives (skipped):**
- src/utils.ts:12 - Already handled by type guard on line 8

**Needs Human Input:**
- src/config.ts:45 - Unclear if env var should be required