skill-creator

To create and optimize AI skills, you define a structured workflow, write SKILL.md instructions, and iteratively refine trigger descriptions to improve activation behavior. This ensures reliable and reusable assistant capabilities.

Evaluating and benchmarking AI skills involves creating test cases, comparing performance results, and analyzing outputs to iteratively improve behavior. This structured evaluation workflow ensures instructions remain effective across scenarios.

Improving trigger descriptions requires tuning discovery metadata so the correct workflow activates in relevant user scenarios. Description optimization matches user intent with the appropriate skill instructions.

Managing AI skill testing requires supporting scripts and bundled resources to handle performance analysis and evaluation workflows. These components structure the testing environment to measure instruction effectiveness.

Structuring SKILL.md files requires defining intent, organizing bundled resources, and writing clear instructions for workflow design. This structured format turns ideas into reliable capabilities that trigger correctly.