.apm/skills/test-verifier/SKILL.md
This skill should be used when the user asks to 'verify tests', 'check coverage', 'detect phantom completions', 'check spec drift', or needs guidance on Phase 8 verification. Also trigger on 'sdd test', 'test traceability', 'verification report', or 'gate criteria'.
npx skillsauth add paulasilvatech/specky Test Verification SpecialistInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Phase 8 systematically verifies that implementation matches specification, achieves required test coverage, detects phantom completions, and identifies spec-code drift.
Execute tests against implementation and generate comprehensive verification report:
All four criteria must pass to advance to Phase 9:
Zero phantom completions. A phantom completion is a test that passes without validating the intended requirement:
Common patterns (detect and fix):
Use mutation testing and requirement traceability to identify phantoms.
Maximum 20% acceptable drift. Drift occurs when:
Drift analysis output:
Total requirements: 150
Implemented as-spec: 130 (86.7%)
Minor drift: 15 (10%)
Major drift: 5 (3.3%)
Drift ratio: 13% ✓ PASS (≤20%)
Systematic approach to identifying false-positive tests:
Link each test to specific SPECIFICATION.md requirement:
Test: test_user_login_with_invalid_password
Requirement: REQ-AUTH-001 (System shall validate password strength)
Coverage: Test validates lowercase-only password rejected
Status: ✓ VALID (tests actual requirement)
Introduce deliberate code changes and verify tests catch them:
If tests pass after mutation, mark as phantom.
Review all mocks and stubs:
Verify test assertions validate actual behavior:
assert response.status == 200 alone → Partial (adds Gate 2)assert response.status == 200 AND response.data.id is not None → Betterassert response.status == 200 AND response.data == expected_user → VALIDPerform bidirectional analysis comparing SPECIFICATION.md to implementation:
For each requirement in SPECIFICATION.md:
For each implementation feature:
## Spec-Code Drift Analysis
### SPECIFICATION.md Coverage: 86.7%
#### Unimplemented Requirements (4)
- REQ-PAYMENT-005: Multi-currency support → STATUS: Deferred to v2
- REQ-ADMIN-002: Bulk user export → STATUS: In progress
- REQ-REPORT-001: Custom reports → STATUS: Not started
- REQ-MOBILE-001: Native mobile app → STATUS: Out of scope
#### Partially Implemented (8)
- REQ-AUTH-001: Password reset → Missing email verification step
- REQ-SEARCH-002: Advanced filters → Partial (date filter missing)
### Implementation Coverage: 95%
#### Undocumented Features (5)
- CSV import from settings menu → Added QA request
- Automatic backup on deploy → Config-driven
- Rate limiting on APIs → Security measure, documented in DESIGN.md
### Drift Ratio: 13% PASS
Produce standardized verification artifact:
# Verification Report
**Date:** [Date]
**Version:** [spec version]
**Prepared by:** [automation/analyst]
## Executive Summary
- Tests executed: [count]
- Pass rate: [%]
- Coverage: [%]
- Drift: [%]
- Gate status: [PASS/FAIL]
## Test Results
### Unit Tests
- Executed: [count]
- Passed: [count]
- Failed: [count]
- Skipped: [count]
### Integration Tests
[similar structure]
### End-to-End Tests
[similar structure]
## Coverage Analysis
[by module with percentages]
## Phantom Completions
- Detected: [count]
- Resolved: [count]
- Status: [VERIFIED CLEAN]
## Spec-Code Drift
[drift analysis with table]
## Gate Criteria Status
- [ ] ≥90% test pass rate
- [ ] ≥80% code coverage
- [ ] Zero phantom completions
- [ ] Drift ≤20%
**Overall Gate Status:** PASS / CONDITIONAL / FAIL
## Recommendations
[Items for Phase 9 release]
/specky:verify --full # Complete verification
/specky:verify --coverage # Coverage report only
/specky:verify --drift # Spec-code drift analysis
/specky:verify --phantoms # Phantom detection only
/specky:verify --gates # Gate criteria check
| Tool | Purpose |
|------|---------|
| sdd_verify_tests | Parse test results, map to REQ-IDs, build coverage report |
| sdd_verify_tasks | Detect phantom completions (tasks done but tests failing) |
| sdd_check_sync | Detect spec-code drift (requirements vs implementation) |
| sdd_validate_ears | Re-validate EARS pattern integrity |
| sdd_get_status | Check pipeline state and current phase |
@test-verifier — Phase 8 agent that calls these tools in sequence. Load this skill as first step.
devops
This skill should be used when the user asks to 'orchestrate pipeline', 'run full pipeline', 'coordinate phases', 'advance phase', or needs guidance on end-to-end pipeline execution. Also trigger on 'specky orchestrate', 'phase transition', 'LGTM gate', 'pipeline status', or 'which agent for this phase'.
tools
This skill should be used when the user asks 'help me with specky', 'what is specky', 'how to use specky', 'specky onboarding', 'get started', or needs guidance on project setup, work modes, or available agents/tools/prompts. Also trigger on '@specky', 'specky help', or 'show me specky commands'.
development
This skill should be used when the user asks about 'spec-driven development', 'SDD pipeline', 'specky', 'pipeline phases', 'EARS notation', 'requirements engineering', 'model routing', or needs guidance on the 10-phase SDD workflow (Init → Discover → Specify → Clarify → Design → Tasks → Analyze → Implement → Verify → Release). Also trigger on 'spec this', 'plan this feature', 'break into tasks', 'quality gate', 'constitution', or 'spec sync'.
development
This skill should be used when generating or formatting SDD artifacts (CONSTITUTION.md, SPECIFICATION.md, DESIGN.md, TASKS.md, VERIFICATION.md, ANALYSIS.md). Also trigger on 'sdd markdown', 'artifact format', or 'spec formatting'.