debug-toolbar: Command for Claude Code

Review Workflow

You are reviewing the implementation of: $ARGUMENTS

Pre-Review Checklist

1. Load PRD from specs/active/{slug}/prd.md
2. Load quality gates from specs/guides/quality-gates.yaml
3. Review all modified files
4. Check test coverage

Critical Rules

1. ALL GATES MUST PASS - No exceptions
2. PATTERN EXTRACTION - Capture new patterns
3. ANTI-PATTERN SCAN - Check for violations
4. DOCUMENTATION UPDATE - Update CLAUDE.md if needed

---

Phase 1: Quality Gate Verification

Run all quality gates:

make test          # All tests pass
make lint          # Zero linting errors
make type-check    # Type checking passes
make fmt-check     # Formatting correct

Verify coverage:

make test-cov
# Check: Modified modules >= 90%

Output: "✓ Phase 1 complete - All quality gates passed"

---

Phase 2: Anti-Pattern Scan

Check for anti-patterns:

1. Optional[T] - Should use T | None

   grep -r "Optional\[" src/ tests/
   

2. Missing future annotations

   for f in $(find src/ -name "*.py"); do
     head -5 "$f" | grep -q "from __future__ import annotations" || echo "Missing: $f"
   done
   

3. Type ignore without code

   grep -r "# type: ignore$" src/ tests/
   

4. Context var cleanup missing

   grep -A5 "@pytest.mark.asyncio" tests/ | grep -v "set_request_context(None)"
   

Fix any violations before proceeding.

Output: "✓ Phase 2 complete - No anti-patterns found"

---

Phase 3: Pattern Extraction

Review new patterns from implementation:

1. Read specs/active/{slug}/tmp/new-patterns.md
2. Identify reusable patterns
3. Extract to specs/guides/patterns/

For each new pattern:

# Pattern: {Name}

## Context
When to use this pattern.

## Implementation
```python
# Code example

Usage

How to apply this pattern.

**Output**: "✓ Phase 3 complete - Patterns extracted"

---