writing-skills

Test-driven AI agent instructions require baseline testing, pressure scenarios, and iterative refinement to validate skill effectiveness before deployment. This methodology ensures reusable instructions reliably guide agents through complex workflows.

Structuring SKILL.md metadata for agent discovery involves writing discovery-focused metadata and organizing supporting resources within the file. This pattern ensures AI agents can effectively discover and follow reusable instructions during complex workflows.

Validating AI skills before deployment requires applying testing practices like baseline testing and pressure scenarios to verify skill effectiveness. This iterative refinement process ensures instructions remain reliable before agents use them in production.

Writing-skills applies to workflows involving authoring new skills, editing existing skills, and verifying skill effectiveness. It provides patterns for structuring SKILL.md files and organizing supporting resources for both creation and modification.

AI skills fail to guide agents when they lack structured SKILL.md metadata, testing practices, and organized supporting resources. Applying test-driven development principles and discovery-focused metadata ensures instructions remain reliable and effective.

Building an internal skill library requires designing, validating, and improving skills that consistently guide agents through complex workflows. This skill provides authoring guidance and testing methodology to ensure library reliability.