X Verify

---
description: Run lint, build, and test verification protocols
---

Use the `ai-engineering-system` skill for additional context about the .ai context system, templates, and workflows.

# AI Engineering System: Verify

Run verification protocols to ensure code quality before considering work complete.

## Steps

1. **Load Verification Config**
   - Read `.ai/TESTING.md` for project-specific verification protocols
   - Check `CLAUDE.md` for any additional verification requirements
   - Fall back to standard commands if no custom config

2. **Run Lint**
   ```bash
   npm run lint

• If errors found: Fix them automatically where possible
• Re-run lint after fixes
• Report any issues that couldn't be auto-fixed

3. Run Build

   npm run build
   

   • If type errors found: Fix them
   • Re-run build after fixes
   • Report any persistent issues

4. Run Tests

   npm run test
   

   • Report test results
   • Do NOT auto-fix failing tests without user approval
   • Summarize which tests failed and why

5. Additional Protocols

   • Run any custom verification from .ai/TESTING.md
   • E2E tests if specified
   • Manual verification checklist items

Iterative Fixing

For lint and build errors:

1. Parse error output
2. Identify fixable issues
3. Apply fixes
4. Re-run verification
5. Repeat until passing or blocked

Maximum 3 iterations per step before reporting to user.

Output Format

## Verification Report

### Lint
- Status: PASS/FAIL
- Issues fixed: [count]
- Remaining issues: [list if any]

### Build
- Status: PASS/FAIL
- Type errors fixed: [count]
- Remaining errors: [list if any]

### Tests
- Status: PASS/FAIL
- Passed: [count]
- Failed: [count]
- Failed tests: [list with reasons]

### Overall
- Status: PASS/FAIL
- Ready for commit: YES/NO

On Failure

If verification fails and cannot be auto-fixed:

1. Report specific failures
2. Suggest fixes if apparent
3. Ask user how to proceed
4. Do NOT mark tasks complete until verification passes