Create, evaluate, and refine AI agent skills with SKILL.md files.
npx skills add https://github.com/mysticfall/alleycat --skill skill-creator-mysticfall
Please help me install this Agent Skill. Skill: skill-creator Source: https://github.com/mysticfall/alleycat/tree/main/.opencode/skills/skill-creator Command: npx skills add https://github.com/mysticfall/alleycat --skill skill-creator-mysticfall
Managing and scaling AI agent capabilities is hard. This Skill Unit provides a repeatable process to create, update, evaluate, and improve AI agent skills, turning ad-hoc experiments into a maintainable workflow.
Define the skill's goal, draft the SKILL.md, run trigger evaluation on realistic prompts, and iterate until you converge on a stable description.
High-intent search queries and answers about installing and using this skill.
To create and test AI skills, define the skill's goal and success criteria, draft the SKILL.md, run trigger evaluations on realistic prompts, and iterate until the description stabilizes. This coordinates utility scripts and references into a repeatable workflow.
The best way to manage prompt engineering iteration is running evaluation loops that refine prompts, metrics, and descriptions. This approach captures intent and turns ad-hoc prompt experiments into a maintainable, testable skill-creation workflow.
To package and distribute AI agent skills, validate your SKILL.md, test prompts, and references, then package them into .skill files. This process ensures your skills are organized and ready for reuse across different environments.
Yes, you can use YAML to organize skill documentation. The workflow relies on PyYAML to manage SKILL.md content and references, ensuring your AI skill metadata and configurations are structured and valid.
Skill trigger evaluation fails when the description, metrics, and prompts are misaligned. You must run iterative improvement loops to refine the SKILL.md triggers and success criteria until the evaluation converges on a stable result.
Yes, utility scripts are required to automate AI skill creation and refinement. They coordinate the generation of SKILL.md, manage test prompts, and execute evaluation loops to produce repeatable, testable Skill Units.
AzrinPutraGuide exploratory conversations to elicit project requirements and generate an intent brief.
billlzzz26Define and package Skill Units with SKILL.md frontmatter and scripts.
breath57Create, test, and refine AI skills using the Anthropic skills specification.
miles990Apply test-driven development cycles to create structured SKILL.md documentation.
