| name | testing-workflow |
| description | Use when someone wants to build a complete, stable test suite from scratch — starting with TDD discipline, moving into coverage strategy, and ending with flaky test elimination. Triggers on: "testing workflow", "테스트 전체 프로세스", "TDD부터 커버리지까지", "full |
| type | workflow |
| theme | engineering |
| scenarios | ["testing workflow 전체 돌려줘","TDD부터 커버리지 전략까지 단계별로 해보자","full test cycle from scratch","CI 테스트 안정화 프로세스 시작","테스트 체계 제대로 잡고 싶어","flaky test 포함해서 전체 테스트 전략 짜줘"] |
| estimated_time | 2-8 hours (full), 30-120 min per step |
| compatibility | {"recommended":["think-tool"],"optional":["sequential-thinking"],"remote_mcp_note":"think-tool은 RED 단계 실패 검증과 커버리지 갭 분석에서 특히 유효합니다. sequential-thinking은 flaky test 트리아지 순서를 강제해 원인 건너뜀을 방지합니다. Claude 설정 → MCP Servers에서 remote SSE 엔드포인트를 추가하세요."} |
Testing Workflow
3-step test quality process: write first → strategize coverage → stabilize CI.
When to Use / When Not to Use
| Use | Skip |
|---|
| Starting a new feature with no tests | Adding a single test to already-covered code |
| Test suite exists but has no coverage strategy | Consistent (non-flaky) test failures — those are bugs |
| CI is failing intermittently due to flaky tests | Quick one-off unit test request |
Workflow Overview
[1] Test-Driven Development
Write failing test → minimal code → refactor
↓
[2] Test Coverage Strategy
Audit coverage gaps → test pyramid → test plan
↓
[3] Flaky Test Elimination
Triage → root cause → fix → CI hardening
Steps
Step 1 — Test-Driven Development
Skill: test-driven-development
Goal: Drive new feature implementation through failing tests first; enforce red-green-refactor discipline
Input: Feature requirements or acceptance criteria
Output: Passing test suite built test-first; implementation with no production code written before a failing test
Skip if: Feature is a throwaway prototype or you are adding tests to existing untested code (go to Step 2)
"Step 1 시작" 또는 "TDD로 이 기능 개발해줘"
Step 2 — Test Coverage Strategy
Skill: test-master
Goal: Audit coverage gaps, define the test pyramid, produce a formal test plan covering unit, integration, E2E, performance, and security tiers
Input: Codebase or module from Step 1 (or existing code if skipping Step 1)
Output: Coverage gap report, test plan with severity ratings, scaffolded test cases for uncovered paths
Skip if: Coverage targets are already met and a current test plan exists
"Step 2 시작" 또는 "테스트 커버리지 분석하고 전략 잡아줘"
Step 3 — Flaky Test Elimination
Skill: flaky-test-analyzer
Goal: Diagnose and fix intermittent failures in CI; apply prevention rules to the whole suite
Input: Failing CI run logs or list of known flaky tests from Step 2
Output: Root cause per flaky test, concrete fix (code + CI config), prevention rule added to suite guidelines
Skip if: All tests pass consistently in CI with no re-run workarounds
"Step 3 시작" 또는 "flaky test 원인 찾고 고쳐줘"
State Tracking
어느 단계에 있는지 알려주면 바로 합류합니다:
- "커버리지부터" → Step 2로 직행
- "flaky test만 봐줘" → Step 3로 직행
- Skip 조건에 해당하면 자동으로 다음 단계로 안내
What Claude Does / What You Do
| Claude | You |
|---|
| Writes failing tests with clear names (Step 1) | Confirm the test tests the right behavior |
| Writes minimal passing implementation (Step 1) | Run tests and verify pass/fail as expected |
| Coverage gap analysis, test plan scaffolding (Step 2) | Provide existing test files and coverage reports |
| Flakiness root cause + fix patterns (Step 3) | Run the failing test 20 times in isolation to confirm |
| CI config recommendations (Step 3) | Wire fixes and configure CI pipeline |
Related Skills
- Individual skills:
develop:test-driven-development, develop:test-master, develop:flaky-test-analyzer
- Before:
develop:dev-quality-workflow (if you want the full dev cycle including architecture and docs)
- After:
develop:performance-profiling-optimization (once the test suite is green and stable)
