writing-skills

To write AI skills with TDD, you apply RED-GREEN-REFACTOR cycles to author and edit SKILL.md documentation, ensuring the skills teach the right thing through pressure scenario testing before deployment.

The best way to test AI skills is by writing pressure scenarios and baseline tests within a test-driven development framework, verifying that the documentation adheres to standards and reliably teaches the intended behavior.

Test-driven development applies to skill documentation by using RED-GREEN-REFACTOR cycles to guide the authoring and editing process, ensuring the documentation meets compliance standards and effectively instructs the AI.

Yes, skill creation requires specific YAML frontmatter to ensure the final AI skill is properly formatted for discovery and adheres to the necessary documentation standards before deployment.

Yes, you can update existing AI skills using a test-driven methodology by applying RED-GREEN-REFACTOR cycles to edit the current documentation, running pressure scenario tests to verify compliance and effectiveness.

You should use a TDD framework for skill authoring when you need to create a new AI skill or update an existing one, ensuring the output is well-documented, reliable, and meets compliance standards before deployment.