stay-green

name	stay-green
description	2-gate TDD development workflow: Gate 1 is Red-Green-Refactor testing, Gate 2 is pre-commit quality checks. Use when implementing features, fixing bugs, or doing any development work. Ensures code is never committed without passing tests and quality checks. Do NOT use for bug-specific debugging (use bug-squashing-methodology).
metadata	{"author":"Geoff","version":"1.0.0"}

Stay Green

Write tests first, then code. Never declare work finished until all checks pass.

Instructions

Gate 1: TDD (Red-Green-Refactor)

1. Red - Write a failing test describing the behavior you want

   ./scripts/test.sh --all  # Should fail
   

2. Green - Write just enough code to make the test pass

   ./scripts/test.sh --all  # Should pass
   

3. Refactor - Clean up while keeping tests green

   ./scripts/test.sh --all  # Should still pass
   

Repeat for each small piece of functionality. Write tests incrementally, not all at once.

Gate 2: Pre-Commit Quality Checks

pre-commit run --all-files

When checks fail: read errors, fix issues, run again. Repeat until all green.

Quality checks include: formatting (Black + isort), linting (Ruff), type checking (MyPy), complexity (<=10 per function), security (Bandit), tests with coverage (>=90%), file hygiene.

Work is DONE when:

1. All tests pass (Gate 1 complete)
2. All pre-commit checks pass (Gate 2 complete)

No exceptions.

After Push: Await the Reviewer (if applicable)

If the repo runs the Claude reviewer GitHub Action, "done" extends past local gates: the latest verdict for HEAD must be LGTM. Don't poll, don't sleep, and don't wait on "CI green" as a proxy — the PR webhook does not deliver CI passes, only comments and CI failures. Use await-claude-review to subscribe and end the turn; the bot's verdict comment wakes the session via <github-webhook-activity>. From there, address-feedback handles the merge gate or the fix loop, including calling mcp__github__unsubscribe_pr_activity once the PR merges or closes.

Examples

Example 1: Adding a New Function

# Gate 1 - Red: Write failing test
def test_calculate_cost_from_impressions():
    result = calculate_cost(impressions=1000, cpm=5.0)
    assert result == 5.0

# Gate 1 - Green: Make it pass
def calculate_cost(impressions: int, cpm: float) -> float:
    return impressions * (cpm / 1000)

# Gate 1 - Refactor: (already clean, move on)
# Gate 2: pre-commit run --all-files -> All passed!

Example 2: Fixing a Formatting Failure

# Gate 2 fails on formatting
$ pre-commit run --all-files
black....Failed

# Auto-fix and re-run
$ ./scripts/format.sh --fix
$ pre-commit run --all-files
# All passed!

Troubleshooting

Error: Coverage below 90%

./scripts/test.sh --all --coverage  # See what's not covered
# Add tests for uncovered lines, then re-run pre-commit

Error: Complexity above 10

./scripts/complexity.sh  # Find complex functions
# Extract helper functions, simplify branching
# Then verify: ./scripts/complexity.sh && pre-commit run --all-files

Error: Type errors from MyPy

./scripts/typecheck.sh  # See specific errors
# Add/fix type annotations
# Then verify: ./scripts/typecheck.sh && pre-commit run --all-files