本文へ移動
cccskills
無料GitHub で公開

tdd-cycle

Execute a comprehensive Test-Driven Development (TDD) workflow with strict red-green-refactor discipline:

インストール方法を見る

含まれるファイル(1)

  • SKILL.md8.8 KB

SKILL.md(原文)

インストールする前に、エージェントに与えられる指示の中身を確認できます。

Execute a comprehensive Test-Driven Development (TDD) workflow with strict red-green-refactor discipline:

[Extended thinking: This workflow enforces test-first development through coordinated agent orchestration. Each phase of the TDD cycle is strictly enforced with fail-first verification, incremental implementation, and continuous refactoring. The workflow supports both single test and test suite approaches with configurable coverage thresholds.]

Configuration

Coverage Thresholds

  • Minimum line coverage: 80%
  • Minimum branch coverage: 75%
  • Critical path coverage: 100%

Refactoring Triggers

  • Cyclomatic complexity > 10
  • Method length > 20 lines
  • Class length > 200 lines
  • Duplicate code blocks > 3 lines

Phase 1: Test Specification and Design

1. Requirements Analysis

  • Do this step directly in Codex CLI (legacy playbook referenced subagent: comprehensive-review::architect-review).
  • Instruction: "Analyze requirements for: (use the user's prompt). Define acceptance criteria, identify edge cases, and create test scenarios. Output a comprehensive test specification."
  • Output: Test specification, acceptance criteria, edge case matrix
  • Validation: Ensure all requirements have corresponding test scenarios

2. Test Architecture Design

  • Do this step directly in Codex CLI (legacy playbook referenced subagent: unit-testing::test-automator).
  • Instruction: "Design test architecture for: (use the user's prompt) based on test specification. Define test structure, fixtures, mocks, and test data strategy. Ensure testability and maintainability."
  • Output: Test architecture, fixture design, mock strategy
  • Validation: Architecture supports isolated, fast, reliable tests

Phase 2: RED - Write Failing Tests

3. Write Unit Tests (Failing)

  • Do this step directly in Codex CLI (legacy playbook referenced subagent: unit-testing::test-automator).
  • Instruction: "Write FAILING unit tests for: (use the user's prompt). Tests must fail initially. Include edge cases, error scenarios, and happy paths. DO NOT implement production code."
  • Output: Failing unit tests, test documentation
  • CRITICAL: Verify all tests fail with expected error messages

4. Verify Test Failure

  • Do this step directly in Codex CLI (legacy playbook referenced subagent: tdd-workflows::code-reviewer).
  • Instruction: "Verify that all tests for: (use the user's prompt) are failing correctly. Ensure failures are for the right reasons (missing implementation, not test errors). Confirm no false positives."
  • Output: Test failure verification report
  • GATE: Do not proceed until all tests fail appropriately

Phase 3: GREEN - Make Tests Pass

5. Minimal Implementation

  • Do this step directly in Codex CLI (legacy playbook referenced subagent: backend-development::backend-architect).
  • Instruction: "Implement MINIMAL code to make tests pass for: (use the user's prompt). Focus only on making tests green. Do not add extra features or optimizations. Keep it simple."
  • Output: Minimal working implementation
  • Constraint: No code beyond what's needed to pass tests

6. Verify Test Success

  • Do this step directly in Codex CLI (legacy playbook referenced subagent: unit-testing::test-automator).
  • Instruction: "Run all tests for: (use the user's prompt) and verify they pass. Check test coverage metrics. Ensure no tests were accidentally broken."
  • Output: Test execution report, coverage metrics
  • GATE: All tests must pass before proceeding

Phase 4: REFACTOR - Improve Code Quality

7. Code Refactoring

  • Do this step directly in Codex CLI (legacy playbook referenced subagent: tdd-workflows::code-reviewer).
  • Instruction: "Refactor implementation for: (use the user's prompt) while keeping tests green. Apply SOLID principles, remove duplication, improve naming, and optimize performance. Run tests after each refactoring."
  • Output: Refactored code, refactoring report
  • Constraint: Tests must remain green throughout

8. Test Refactoring

  • Do this step directly in Codex CLI (legacy playbook referenced subagent: unit-testing::test-automator).
  • Instruction: "Refactor tests for: (use the user's prompt). Remove test duplication, improve test names, extract common fixtures, and enhance test readability. Ensure tests still provide same coverage."
  • Output: Refactored tests, improved test structure
  • Validation: Coverage metrics unchanged or improved

Phase 5: Integration and System Tests

9. Write Integration Tests (Failing First)

  • Do this step directly in Codex CLI (legacy playbook referenced subagent: unit-testing::test-automator).
  • Instruction: "Write FAILING integration tests for: (use the user's prompt). Test component interactions, API contracts, and data flow. Tests must fail initially."
  • Output: Failing integration tests
  • Validation: Tests fail due to missing integration logic

10. Implement Integration

  • Do this step directly in Codex CLI (legacy playbook referenced subagent: backend-development::backend-architect).
  • Instruction: "Implement integration code for: (use the user's prompt) to make integration tests pass. Focus on component interaction and data flow."
  • Output: Integration implementation
  • Validation: All integration tests pass

Phase 6: Continuous Improvement Cycle

11. Performance and Edge Case Tests

  • Do this step directly in Codex CLI (legacy playbook referenced subagent: unit-testing::test-automator).
  • Instruction: "Add performance tests and additional edge case tests for: (use the user's prompt). Include stress tests, boundary tests, and error recovery tests."
  • Output: Extended test suite
  • Metric: Increased test coverage and scenario coverage

12. Final Code Review

  • Do this step directly in Codex CLI (legacy playbook referenced subagent: comprehensive-review::architect-review).
  • Instruction: "Perform comprehensive review of: (use the user's prompt). Verify TDD process was followed, check code quality, test quality, and coverage. Suggest improvements."
  • Output: Review report, improvement suggestions
  • Action: Implement critical suggestions while maintaining green tests

Incremental Development Mode

For test-by-test development:

  1. Write ONE failing test
  2. Make ONLY that test pass
  3. Refactor if needed
  4. Repeat for next test

Use this approach by adding --incremental flag to focus on one test at a time.

Test Suite Mode

For comprehensive test suite development:

  1. Write ALL tests for a feature/module (failing)
  2. Implement code to pass ALL tests
  3. Refactor entire module
  4. Add integration tests

Use this approach by adding --suite flag for batch test development.

Validation Checkpoints

RED Phase Validation

  • All tests written before implementation
  • All tests fail with meaningful error messages
  • Test failures are due to missing implementation
  • No test passes accidentally

GREEN Phase Validation

  • All tests pass
  • No extra code beyond test requirements
  • Coverage meets minimum thresholds
  • No test was modified to make it pass

REFACTOR Phase Validation

  • All tests still pass after refactoring
  • Code complexity reduced
  • Duplication eliminated
  • Performance improved or maintained
  • Test readability improved

Coverage Reports

Generate coverage reports after each phase:

  • Line coverage
  • Branch coverage
  • Function coverage
  • Statement coverage

Failure Recovery

If TDD discipline is broken:

  1. STOP immediately
  2. Identify which phase was violated
  3. Rollback to last valid state
  4. Resume from correct phase
  5. Document lesson learned

TDD Metrics Tracking

Track and report:

  • Time in each phase (Red/Green/Refactor)
  • Number of test-implementation cycles
  • Coverage progression
  • Refactoring frequency
  • Defect escape rate

Anti-Patterns to Avoid

  • Writing implementation before tests
  • Writing tests that already pass
  • Skipping the refactor phase
  • Writing multiple features without tests
  • Modifying tests to make them pass
  • Ignoring failing tests
  • Writing tests after implementation

Success Criteria

  • 100% of code written test-first
  • All tests pass continuously
  • Coverage exceeds thresholds
  • Code complexity within limits
  • Zero defects in covered code
  • Clear test documentation
  • Fast test execution (< 5 seconds for unit tests)

Notes

  • Enforce strict RED-GREEN-REFACTOR discipline
  • Each phase must be completed before moving to next
  • Tests are the specification
  • If a test is hard to write, the design needs improvement
  • Refactoring is NOT optional
  • Keep test execution fast
  • Tests should be independent and isolated

TDD implementation for: (use the user's prompt)

Output Format

<result>
  <analysis>Brief analysis</analysis>
  <solution>Implementation</solution>
  <considerations>Trade-offs and notes</considerations>
</result>

レビュー

まだレビューはありません。使ってみた感想をお寄せください。

同じリポジトリのスキル

概要と使いどころ

Audit UI code for WCAG compliance

日本語の概要は準備中です。原文の説明を表示しています。

itsimonfredlingjack/codex-dev-plugin22026年2月5日 更新

Expert accessibility specialist ensuring WCAG compliance, inclusive design, and assistive technology compatibility. Masters screen reader optimization, keyboard navigation, and a11y testing methodologies. Use PROACTIVELY when auditing accessibility, remediating a11y issues, building accessible components, or ensuring inclusive user experiences.

日本語の概要は準備中です。原文の説明を表示しています。

itsimonfredlingjack/codex-dev-plugin22026年2月5日 更新

Skill for accessibility-patterns tasks.

日本語の概要は準備中です。原文の説明を表示しています。

itsimonfredlingjack/codex-dev-plugin22026年2月5日 更新

Build AI assistant application with NLU, dialog management, and integrations

日本語の概要は準備中です。原文の説明を表示しています。

itsimonfredlingjack/codex-dev-plugin22026年2月5日 更新

Build production-ready LLM applications, advanced RAG systems, and intelligent agents. Implements vector search, multimodal AI, agent orchestration, and enterprise AI integrations. Use PROACTIVELY for LLM features, chatbots, AI agents, or AI-powered applications.

日本語の概要は準備中です。原文の説明を表示しています。

itsimonfredlingjack/codex-dev-plugin22026年2月5日 更新

ai-review

無料

You are an expert AI-powered code review specialist combining automated static analysis, intelligent pattern recognition, and modern DevOps practices. Leverage AI tools (GitHub Copilot, Qodo, GPT-5, Claude 4.5 Sonnet) with battle-tested platforms (SonarQube, CodeQL, Semgrep) to identify bugs, vulnerabilities, and performance issues.

日本語の概要は準備中です。原文の説明を表示しています。

itsimonfredlingjack/codex-dev-plugin22026年2月5日 更新

itsimonfredlingjack のスキルをすべて見る

このスキルの問題を報告する