Verify

---
name: verify
description: Verify code changes work by running them. Proves through execution, not
  just inspection.
---

# /verify — Prove Code Works

$ARGUMENTS

---

## 🔴 CRITICAL RULES

1. **Load verify-changes skill** — Read `.agent/skills/verify-changes/SKILL.md` first
2. **Execute, don't inspect** — Run the code, don't just read it
3. **Report evidence** — Show actual output, not assumptions
4. **Cover edge cases** — Test error paths, not just happy path

---

## Task

Use the `verify-changes` skill to prove code works:

```
CONTEXT:
- What to verify: $ARGUMENTS
- If empty: verify the most recent code changes in this session

WORKFLOW:
1. IDENTIFY what changed (files, functions, behavior)
2. DETERMINE verification method (build, test, run, curl)
3. EXECUTE verification commands
4. REPORT evidence of success or failure
5. FLAG anything that couldn't be verified automatically

RULES:
1. Follow verify-changes/SKILL.md protocol
2. "It should work" is NOT verification — run it
3. Test error paths, not just success paths
4. Report with actual command output as evidence
```

---

## Expected Output

```
## Verification Report

### Changes Verified
- [file/change 1]: ✅ Pass
- [file/change 2]: ✅ Pass

### Evidence
- Build: ✅ Compiled without errors
- Tests: ✅ [N]/[N] passing
- Runtime: ✅ [specific verification result]

### Not Verified
- [anything that needs manual testing]
```

---

## Usage Examples

```
/verify
/verify the login endpoint handles expired tokens
/verify build passes after refactoring
/verify the new component renders correctly
```

---
*Generated by [Agent Bridge](https://github.com/HaoNgo232/agent-bridge)*