ces-ltd/lumi/skill-verify

BEFORE claiming any status or expressing satisfaction:
 
1. IDENTIFY → What command proves this claim?
2. RUN      → Execute the FULL command (fresh, complete)
3. READ     → Full output, check exit code, count failures
4. VERIFY   → Does output confirm the claim?
               - If NO: State actual status with evidence
               - If YES: State claim WITH evidence
5. CLAIM    → ONLY THEN make the claim
 
Skip any step = lying, not verifying

Running tests...
$ npm test
 
  ✓ user.create() saves to database (45ms)
  ✓ user.create() validates email (12ms)
  ✓ user.create() hashes password (23ms)
 
Tests: 3 passed, 3 total
Time: 0.8s
 
All 3 tests pass. Ready for review.

I've implemented the user creation feature. It should work now.
The tests should pass.

1. Write test:
   test('rejects empty email', () => { ... })
 
2. Run test (should FAIL):
   $ npm test
   FAIL: expected 'Email required', got undefined
   ✓ Test fails as expected
 
3. Implement fix
 
4. Run test (should PASS):
   $ npm test
   PASS: 1/1
   ✓ Test passes
 
5. Revert fix temporarily:
   $ git stash
   $ npm test
   FAIL: expected 'Email required', got undefined
   ✓ Confirms test catches the bug
 
6. Restore fix:
   $ git stash pop
   $ npm test
   PASS: 1/1
 
Red-green cycle verified. Regression test is valid.

# Run and capture output
npm test                  # JavaScript/TypeScript
pytest                    # Python
cargo test                # Rust
go test ./...             # Go
 
# Expected: "X passed, 0 failed"

npm run build             # Node.js
cargo build --release     # Rust
go build ./...            # Go
 
# Expected: Exit code 0, no errors

npm run lint              # ESLint
ruff check .              # Python
cargo clippy              # Rust
 
# Expected: 0 errors (warnings OK)

npm run typecheck         # TypeScript
mypy .                    # Python
cargo check               # Rust
 
# Expected: 0 errors

# Before moving from tangle → ink
${HOME}/.claude-octopus/plugin/scripts/orchestrate.sh preflight
 
# Verifies:
# - All agents completed
# - Quality gate passed
# - No errors in logs

# Full verification before PR
npm test && npm run lint && npm run build
 
# Then create PR with evidence
gh pr create --body "$(cat <<'EOF'
## Verification
- Tests: 42/42 passing
- Lint: 0 errors
- Build: Success
EOF
)"

Claiming success → Verification evidence exists in this message
Otherwise → Not verified