Create and iteratively improve AI agent skills through structured evaluation workflows.
npx skills add https://github.com/cline/skills --skill skill-creator-cline
Please help me install this Agent Skill. Skill: skill-creator Source: https://github.com/cline/skills/tree/main/skills/skill-creator Command: npx skills add https://github.com/cline/skills --skill skill-creator-cline
Building effective AI agent skills requires more than just writing instructions—it demands systematic testing, measurement, and iterative refinement to ensure reliable triggering and consistent performance across diverse user queries.
Use the skill-creator skill to build a new skill for extracting invoice data from PDFs by describing the workflow and test cases you want to validate.
High-intent search queries and answers about installing and using this skill.
AI agent skills are improved through structured evaluation workflows that apply quantitative benchmarking, blind comparison testing, and description optimization for trigger accuracy. This systematic approach measures reliable performance across diverse user queries.
Optimize AI agent skill descriptions for trigger accuracy by using train/test splits with Claude-powered iteration loops. This automated description improvement loop refines how reliably skills activate for intended queries.
Package and distribute AI workflows by validating and compressing them into distributable .skill files. This packaging process enables installation across multiple AI agents for standardized workflow deployment.
Python scripts are required for AI agent evaluation to handle eval aggregation, HTML report generation, and Claude Code CLI integration. These scripts automate the iterative testing and description improvement loops.
PyYAML is required as a dependency for AI workflow skill creation and editing. It supports the structured workflows needed to draft new skills from user intent or improve existing ones.
yanochka11Scan agent skill definitions and scripts for malicious patterns.
jamesogunsanAutomate the OpenCode skill lifecycle from evaluation loops to packaging.
qiushibangCreate, modify, and improve AI agent skills through eval-driven workflows.
mymx2Draft, test, and benchmark AI skill instructions with automated evaluation loops.
