nw-review-workflow

# Agent Review Workflow ## V2 Validation Checklist (11 Points) Run against every agent under review. Each item pass/fail. 1. **Frontmatter format**: `---` delimited YAML with `name` and `description` 2. **Size compliance**: Under 400 lines; domain knowledge in Skills 3. **Divergence-only**: Only behaviors diverging from Claude defaults 4. **Calm language**: No "CRITICAL", "MANDATORY", "ABSOLUTE" 5. **Examples present**: 3-5 canonical for critical/subtle behaviors 6. **Least privilege tools**: Minimum needed in frontmatter 7. **maxTurns set**: Present in frontmatter 8. **Platform safety**: Via frontmatter/hooks, not prose 9. **Affirmative phrasing**: "Do X" not "Don't do Y" 10. **Consistent terminology**: One term per concept 11. **Clear delegation**: Description states when to delegate ## Scoring Methodology ### Per-Dimension Scoring For each of 7 critique dimensions (from critique-dimensions skill): - **Pass**: All checks satisfied - **Fail**: One or more checks not satisfied ### Verdict Logic ``` IF any high-severity dimension fails: verdict = "revisions_needed" ELIF count(medium-severity failures) >= 3: verdict = "revisions_needed" ELSE: verdict = "approved" ``` High-severity: template_compliance, size_and_focus, safety_implementation, priority_validation Medium-severity: divergence_quality, language_and_tone, examples_quality ### Evidence Requirements Every finding includes: **Dimension** (which of 7) | **Severity** (high/medium/low) | **Finding** (observ...