| name | Code Review |
| trigger | /review-code |
| description | Code-health audit — complexity, duplication, error handling, type invariants, patterns, naming, comment accuracy, dead/unutilized code. Four modes: --review (full pass), --debt (tech-debt catalog), --consolidate (DRY + error-handling consolidation), --patterns (cross-codebase consistency). NOT for security vulns (/security) or performance profiling (/perf). |
| agent | code-reviewer |
| arguments | [{"name":"target","description":"File, directory, or git diff to review (optional, defaults to full codebase)","required":false},{"name":"--review","description":"Default — full 7-dimension health pass → docs/reviews/CODE_REVIEW_<date>.md","required":false},{"name":"--debt","description":"Tech-debt catalog sorted by leverage → docs/reviews/TECH_DEBT_<date>.md","required":false},{"name":"--consolidate","description":"DRY + error-handling consolidation proposals → docs/reviews/CONSOLIDATION_<date>.md","required":false},{"name":"--patterns","description":"Cross-codebase pattern consistency audit → docs/reviews/PATTERNS_<date>.md","required":false}] |
Triggers the code-reviewer subagent.
Reviews code for code health — maintainability, patterns, tech debt, complexity, duplication, error handling, type invariants, naming, and comment accuracy. Distinct from security audit (vulnerabilities) and performance profiling (use /security and /perf).
The 8 dimensions scored on every --review:
- Complexity (function/file length, nesting, cyclomatic)
- Duplication / DRY (copy-paste ratio, missing abstractions)
- Error Handling (silent failures, broad catches, missing context)
- Type Safety & Invariants (illegal states unrepresentable)
- Pattern Consistency (consistency with codebase idioms)
- Naming Quality (intent-revealing, booleans-as-questions)
- Comment Accuracy (comments match code behavior)
- Dead / Unutilized Code (stubs, never-called functions, unused exports, orphan files, disconnected pipelines)
Outputs:
--review → docs/reviews/CODE_REVIEW_<date>.md (Health Dashboard + findings + verdict)--debt → docs/reviews/TECH_DEBT_<date>.md (prioritized debt register)--consolidate → docs/reviews/CONSOLIDATION_<date>.md (DRY + error-handling refactor proposals)--patterns → docs/reviews/PATTERNS_<date>.md (cross-codebase drift audit)
Reference: references/code-health-checklist.md (read at start of every invocation). Confidence-scored findings (suppress <75), verbatim code quotes, asymmetric gate-loop (< 5 = fail, ≥ 7 = pass).
