skill-creator

To build and optimize AI skills, you define the intent, write the SKILL.md file, and structure the skill directories. This process facilitates the end-to-end lifecycle from initial concept to optimized deployment using iterative testing.

Prompt engineering evaluation works by running automated evaluations and analyzing results to iteratively refine AI skills. It integrates with LLM-based analysis agents to benchmark performance and improve triggering accuracy.

The best way to refine AI skill descriptions is through automated description optimization. This tunes the text for improved discoverability, ensuring Claude or other LLMs trigger the correct skill effectively.

You need automated evaluations to test AI skills because they support iterative improvement. By running test cases and analyzing results, you can refine the skill logic based on concrete LLM feedback.

Yes, you can use this to create a new skill for extracting action items from meeting transcripts. It guides you through writing the SKILL.md, creating test cases, running evaluations, and optimizing the final description.