Test Watch

# Test Watch & Fix

Run tests, monitor results, and auto-fix failures.

## Arguments
$ARGUMENTS - Optional: specific test file or pattern to watch

## Instructions

1. **Detect test framework:**
   - Node: Look for jest, vitest, mocha in package.json
   - Python: pytest, unittest
   - Rust: cargo test
   - Go: go test

2. **Run tests:**
   ```bash
   # Node/npm
   npm test -- --watchAll=false

   # Python
   pytest -v

   # Rust
   cargo test

   # Go
   go test ./...

3. Parse results and categorize:

   • ✅ Passing tests
   • ❌ Failing tests (with error details)
   • ⏭️ Skipped tests

4. For each failure:

   • Read the failing test file
   • Read the source file being tested
   • Analyze the error message
   • Determine if it's a test bug or source bug

5. Auto-fix mode:

   • Ask: "Found {N} failing tests. Auto-fix?"
   • If yes, spawn debugger agent to fix each failure
   • Re-run tests after each fix
   • Continue until all pass or user stops

6. Store results in memory:

   mcp__claude-flow__memory_usage(
     action="store",
     namespace="results",
     key="test-run_{project}_{timestamp}",
     value="{pass_count}/{total} passed, failures: [...]"
   )
   

7. Output:

   ## Test Results: {project}
   
   ✅ {pass_count} passed
   ❌ {fail_count} failed
   ⏭️ {skip_count} skipped
   
   ### Failures
   1. {test_name} - {brief error}
      → Suggested fix: {suggestion}
   
   **Action:** [Fix All] [Fix One-by-One] [Skip]