/verify - Verification Before Completion

Evidence before claims. Always.

Target: $ARGUMENTS

The Iron Law

NO COMPLETION CLAIMS WITHOUT FRESH VERIFICATION EVIDENCE

If you haven't run the verification command in THIS message, you cannot claim it passes.

Violating the letter of this rule is violating the spirit of this rule.

When to Use

• Before committing code
• Before creating a pull request
• Before marking a task as complete
• Before moving to the next task in a plan
• After fixing a bug (prove it's actually fixed)
• NOT for: research or exploration tasks

The Gate Function

BEFORE claiming any status:

1. IDENTIFY - What command proves this claim?
2. RUN - Execute the FULL command (fresh, complete, not cached)
3. READ - Full output, check exit code, count failures
4. VERIFY - Does output ACTUALLY confirm the claim?
   → If NO: State actual status with evidence
   → If YES: State claim WITH the evidence
5. ONLY THEN - Make the claim

Skip any step = lying, not verifying.

Verification Matrix

Operations

/verify test

npm test 2>&1
# Read FULL output
# Count: X passed, Y failed, Z skipped
# Exit code: 0 or non-zero

/verify build

npm run build 2>&1
# Check: exit code 0?
# Check: no error output?

/verify lint

npm run lint 2>&1
# Count: 0 errors, 0 warnings

/verify types

npx tsc --noEmit 2>&1
# Count: 0 errors

/verify all

Run ALL of the above in sequence. Stop at first failure.

/verify (no args)

Auto-detect what to verify based on package.json scripts.

Output

──── /verify ────
Command: npm test
Exit code: 0
Result: 42/42 tests passed, 0 failed

VERIFIED ✓ - All tests pass (evidence above)

Or:

──── /verify ────
Command: npm test
Exit code: 1
Result: 40/42 tests passed, 2 failed

FAILED ✗ - 2 tests failing:
  - src/auth.test.ts: "should validate token" - Expected true, got false
  - src/user.test.ts: "should create user" - Timeout after 5000ms

DO NOT claim completion. Fix failures first.

Red Flags - STOP

If you catch yourself thinking:

• "Should work now"
• "I'm confident this is correct"
• "Just this once I can skip verification"
• "Lint passed, so build will too"
• "I only changed one line"
• "The agent said it succeeded"
• "I'll verify after committing"
• "It's obvious this works"

ALL of these mean: RUN THE COMMAND. READ THE OUTPUT.

Rationalization Prevention

Integration

This skill is a cross-cutting concern. It applies to:

• @builder - Before claiming module is complete
• @executor - Before marking task checkpoint
• @tdd - Before claiming GREEN phase
• @reviewer - Before giving verdict
• @debugger - Before claiming bug is fixed
• @migrator - Before claiming migration is complete

Rules

1. Evidence before claims - Non-negotiable
2. Fresh runs only - Never cite previous runs
3. Full output - Don't truncate or summarize
4. Exit codes matter - 0 = success, anything else = failure
5. Read warnings - Warnings can indicate real problems
6. One claim per verification - Don't bundle claims