guanyang/antigravity-skills

Overview

This skill encodes a disciplined Test-Driven Development (TDD) workflow to use before implementing features or bug fixes. It enforces writing a failing test first, implementing the minimal code to pass, and refactoring while keeping tests green. The aim is higher confidence, clearer design, and repeatable verification for every change.

How this skill works

The skill guides you through the Red-Green-Refactor loop: write one minimal failing test (RED), verify it fails for the expected reason, implement just enough code to make the test pass (GREEN), and then refactor while ensuring all tests remain green. It includes concrete checklists, anti-pattern warnings, and patterns for writing clear, behavior-focused tests rather than implementation-focused ones.

When to use it

• Always when adding new features
• Whenever fixing a bug
• During refactoring that changes behavior
• When changing public APIs or behavior
• Before committing changes that affect logic

Best practices

• Write one behavior-focused test at a time with a clear name
• Verify each new test fails for the intended reason before implementing
• Implement the smallest possible change to make the test pass
• After green, refactor for clarity without changing behavior and keep tests passing
• Avoid over-engineering: add features only when a test requires them
• Delete exploratory code and re-implement from tests to avoid bias

Example use cases

• Add a retry mechanism: write a test asserting three attempts, see it fail, implement minimal retry logic
• Bug fix: reproduce the bug with a failing test, fix code so test passes and prevents regression
• Refactor a module: add tests for current behavior first, then refactor while maintaining green tests
• Design new API: write tests for the desired API behavior before implementing methods
• Validate input rules: write failing tests for edge cases (empty or invalid inputs) then implement validation