"writing-plans vs TDD: Which AI Coding Methodology Fits Your Workflow?"

writing-plans emphasizes "plan first," TDD emphasizes "test first." They're not mutually exclusive — you can use both.

What This Is

We read both skills' SKILL.md files and compared their workflows and use cases.

What This Is Not: We did not run the same task through both skills and compare outputs. Our conclusions are based on SKILL.md analysis, not hands-on testing.

Key Differences

writing-plans Workflow

Per the SKILL.md:

1. AI asks clarifying questions first
2. Outputs a detailed implementation plan: which files to change, what to change, how to test
3. Each task in the plan is a 2-5 minute step
4. After you approve, AI executes step by step

TDD Workflow

Per the SKILL.md:

1. Write the test first
2. Run test, confirm it fails (Red)
3. Write minimal code to pass (Green)
4. Refactor if needed
5. Commit

When to Use Which?

Use writing-plans for:

• ✅ New features (involving multiple files)
• ✅ Tasks that need thinking before coding
• ✅ Team collaboration (needs documentation)
• ✅ Unfamiliar domains

Use TDD for:

• ✅ Bug fixes (write reproduction test first)
• ✅ Functions with clear input/output
• ✅ API development
• ✅ Refactoring (tests as safety net)

Use both together:

Large feature: writing-plans for planning → TDD for each task
Small feature: TDD directly

Our Opinion (Editorial, Not Tested)

If you can only pick one: pick writing-plans. Extra planning time might save more rework time. But this is our judgment, not a tested conclusion.

If you can install both: do it. They complement each other.

Install

# writing-plans
git clone --depth 1 https://github.com/obra/superpowers.git && cp superpowers/skills/writing-plans ~/.claude/skills/writing-plans

# TDD
git clone --depth 1 https://github.com/mattpocock/skills.git && cp skills/skills/tdd ~/.claude/skills/tdd

Based on SKILL.md analysis, not hands-on comparison. Your results may vary.