code-review

name	code-review
description	Three-pass code review checking prerequisites, code quality, and test/eval coverage
argument-hint	[workitem-path]
allowed-tools	Read, Glob, Grep, Bash

Code Review

Review code changes for $ARGUMENTS (or current branch changes if no argument).

Pass 1 — Prerequisites & Architecture

• Workitem exists with APPROVED design
• All tasks marked [x] complete in tasks.md
• Layer separation: tools in mortgage_concierge/tools/, models in sub_agents/*/models.py
• No circular imports (check with grep for cross-layer imports)
• State contracts match constants.py

Pass 2 — Code Quality & Security

• Run radon cc -s -n C on modified files — flag CC > 5
• Check for hardcoded secrets, API keys, credentials
• OWASP basics: no eval/exec, no unsafe deserialization, no SQL injection patterns
• Import order: stdlib → third-party → local
• Docstrings on public functions
• Pydantic models have Field descriptions

Pass 3 — Test & Eval Coverage

• For each [TDD] task: verify test exists in tests/unit/ or tests/integration/
• For each [Eval-DD] task: verify eval case exists in tests/eval/data/
• Run uv run python -m pytest tests/unit/ --tb=short — report results
• Check eval thresholds in test_config.json have not been lowered
• New tools must have at least one eval case

Verdict

• LGTM: All passes succeed
• LGTM_WITH_NITS: Minor issues that don't block merge
• REQUEST_CHANGES: Blocking issues found — list each with severity