agentleFS
excattexcatt/superclaude-plusplus/skills/tdd/SKILL.md

tdd

Skill14 starsChanged 37 days agoMITView on GitHub

Test-Driven Development with enforced RED-GREEN-REFACTOR cycle and Git checkpoints per phase. Use for new testable features and bug fixes (reproduction test first) — not for config, docs, or renames.

194 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use capture-learning-tools, design-frontend-dev and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • The-Vibe-CompanySkill

    capture-learning-tools

    Turn technical conversations, bugs, CI failures, review comments,

    Also in 5 projects that use this

  • The-Vibe-CompanySkill

    design-frontend-dev

    Clean Impeccable frontend design skill. Use when the user wants to

    Also in 5 projects that use this

  • tristanphvnSkill

    ketchup-video

    Catch the user up like /keel:ketchup, with a short silent video of slides for what happened and the waiting calls as text under it. Use for /keel:ketchup-video.

    Also in 14 projects that use this

  • jaydubya818Skill

    figure-it-out

    Use when novel or multi-phase work needs an auditable workflow and no narrower playbook fits.

    Also in 54 projects that use this

  • jaydubya818Skill

    principle-laziness-protocol

    Apply when a refactor is adding wrappers or abstractions instead of simplifying.

    Also in 45 projects that use this

  • +2 more from The-Vibe-Company
  • +5 more from jaydubya818

The file

Show its rules and text

We split this skill into the rules it gives an agent

15 rules. 0 are common, 15 are distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 186 edited versions we read keep these.

  • 15

    Distinctive

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Do NOT use for: config changes, documentation, simple renames, UI-only changes

    line 18
  2. Distinctive:

    Goal: Write a test that describes the desired behavior.

    line 53
  3. Distinctive:

    Write the test FIRST — before any production code

    line 56
  4. Distinctive:

    Test should be minimal: one behavior per test

    line 57
  5. Distinctive:

    Run the test suite and confirm failure

    line 58
  6. Distinctive:

    Test fails with assertion error (expected behavior not implemented)

    line 66
  7. Distinctive:

    Test fails with compile/import error (function/class doesn't exist yet)

    line 67
  8. Distinctive:

    Test fails because of typo in test code

    line 70
  9. Distinctive:

    Test fails because test framework misconfigured

    line 71
  10. Distinctive:

    Goal: Write the MINIMUM production code to make the test pass.

    line 80
  11. Distinctive:

    Only write code that the failing test requires

    line 83
  12. Distinctive:

    Run tests after EVERY change — if any test breaks, revert immediately

    line 106
  13. Distinctive:

    All requirements met? → Run full test suite → Report coverage

    line 117
  14. Distinctive:

    After all cycles complete:

    line 121
  15. Distinctive:

    Test-to-test dependencies (each test must be independent)

    line 140

“Common” is measured: kept by at least 6 in 10 of the 186 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

When to Use

Use proactively when:

  • New feature implementation with testable behavior
  • Bug fix (write reproduction test first)
  • User explicitly requests TDD approach

Triggers: TDD, test-driven, test first, red green refactor, write test first, failing test, reproduction test

Do NOT use for: config changes, documentation, simple renames, UI-only changes

Dynamic Context

Recent TDD commits: !git log --oneline -5 --grep="red:\|green:\|refactor:" 2>/dev/null || echo "No TDD commits found"

Test infrastructure: !ls tests/ __tests__/ test/ 2>/dev/null | head -10 || echo "No test directory found"

TDD Workflow Skill

Purpose

Enforce disciplined RED-GREEN-REFACTOR cycle with Git checkpoints to prevent skipping steps.

Usage

/tdd "user login validation"      # new feature
/tdd "fix: timeout on large files" # bug reproduction
/tdd                               # prompt for description

The Cycle

RED → GREEN → REFACTOR → (repeat)
 │      │        │
 │      │        └─ git commit "refactor: ..."
 │      └─ git commit "green: ..."
 └─ git commit "red: ..."

Phase 1: RED (Write Failing Test)

Goal: Write a test that describes the desired behavior. It MUST fail.

Rules

  • Write the test FIRST — before any production code
  • Test should be minimal: one behavior per test
  • Run the test suite and confirm failure

RED Gate (CRITICAL)

DO NOT edit production code until RED state is confirmed.

Acceptable RED states:

  • Test fails with assertion error (expected behavior not implemented)
  • Test fails with compile/import error (function/class doesn't exist yet)

NOT acceptable:

  • Test fails because of typo in test code
  • Test fails because test framework misconfigured

Checkpoint

git add -A && git commit -m "red: [describe expected behavior]"

Phase 2: GREEN (Make It Pass)

Goal: Write the MINIMUM production code to make the test pass. Nothing more.

Rules

  • Only write code that the failing test requires
  • No premature optimization
  • No extra features
  • "Fake it till you make it" is valid (hardcoded returns → generalize later)

GREEN Gate

ALL tests must pass. No new test failures introduced.

Checkpoint

git add -A && git commit -m "green: [describe what was implemented]"

Phase 3: REFACTOR (Clean Up)

Goal: Improve code quality without changing behavior. Tests stay green throughout.

Rules

  • Extract duplications
  • Rename for clarity
  • Simplify logic
  • Run tests after EVERY change — if any test breaks, revert immediately

Checkpoint

git add -A && git commit -m "refactor: [describe improvement]"

Cycle Completion

After REFACTOR, evaluate:

  • More behaviors to implement? → Start new RED phase
  • All requirements met? → Run full test suite → Report coverage

Coverage Gate

After all cycles complete:

# JavaScript/TypeScript
pnpm test -- --coverage

# Python
pytest --cov --cov-report=term-missing

Minimum threshold: Report actual coverage. Flag untested paths. Do not set arbitrary percentage targets unless the project defines one.

Anti-patterns — Must Avoid

  • Writing production code before test (defeats TDD purpose)
  • Writing test that already passes (not RED)
  • Skipping REFACTOR phase ("it works, ship it")
  • Testing implementation details instead of behavior
  • Brittle selectors (CSS classes, DOM structure) in tests
  • Test-to-test dependencies (each test must be independent)
  • Large RED→GREEN jumps (write too much code at once)

Integration with Verification Iron Law

TDD checkpoints feed directly into the Verification Iron Law:

  • RED commit = evidence that test correctly fails
  • GREEN commit = evidence that implementation works
  • REFACTOR commit = evidence that tests remain green after cleanup

Each commit is a verification artifact.

Output Format

## TDD Cycle 1: [behavior]

### RED
- Test: `test/auth.test.ts` — `should reject empty password`
- Status: FAIL ✓ (assertion error: expected rejection)
- Commit: `red: reject empty password validation`

### GREEN
- Implementation: `src/auth.ts:validatePassword()`
- Status: ALL PASS ✓
- Commit: `green: implement password validation`

### REFACTOR
- Change: extracted validation rules to constants
- Status: ALL PASS ✓
- Commit: `refactor: extract validation constants`

## Coverage: 87% (src/auth.ts: 100%)

Related Skills

  • /pytest-runner - Python test execution
  • /verify - Post-implementation verification
  • /confidence-check - Pre-implementation assessment

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.