eval-skills - Evaluate and select AI agent skills
Provides a framework-agnostic toolkit to discover, scaffold, evaluate, select, compare, and report on AI agent skills.
Tags
Updated: 2026-10-02Capabilities
Typical Inputs
Typical Outputs
What this skill does
- Find skills by metadata
- Create skill skeletons
- Run benchmark evaluations
- Select and rank skills
- Run evaluation pipelines
- Convert and compare reports
- Initialize evaluation projects
- Manage configuration values
Inputs
- Skill definition files
- Benchmark task files
- Custom task files
- Evaluation report files
- Selection strategy files
- Project configuration files
- Command-line options
Outputs
- Evaluation reports
- Selected skills files
- Generated skill scaffolds
- Project scaffold files
- Converted reports
- Report diff files
- Process exit status