Overview
agentic-eval is an open-source workflow skill for Claude Code and compatible agents, published by github. Its author describes it as: “Patterns and techniques for evaluating and improving AI agent outputs. Use this skill when: - Implementing self-critique and reflection loops - Building evaluator-optimizer pipelines for quality-critical generation -…”. The project has 38k stars on GitHub and is available under the MIT license. Add it to your setup with `git clone https://github.com/github/awesome-copilot ~/.claude/skills/agentic-eval`.
What agentic-eval does
Patterns for self-improvement through iterative evaluation and refinement.
Best for
What’s inside
The SKILL.md for agentic-eval is organised into these sections:
- Overview
- When to Use
- Pattern 1: Basic Reflection
- Pattern 2: Evaluator-Optimizer
- Pattern 3: Code-Specific Reflection
- Evaluation Strategies
- Outcome-Based
- LLM-as-Judge
- Rubric-Based
- Best Practices
- Quick Start Checklist
Installation
Add agentic-eval to your agent with:
git clone https://github.com/github/awesome-copilot ~/.claude/skills/agentic-evalReview a skill’s source before installing it — skills can instruct your agent to run commands. This command comes from the skill’s public repository, which is the source of truth for setup steps. New to skills? Read how to install Claude Code skills.
Supported agents
- GitHub Copilot — declared in the repository topics
License
agentic-eval is published under the MIT license.
Strengths
- Clear MIT license — safe to read and adapt within its terms
- Ships in github/awesome-copilot, an established project with 37,982 GitHub stars
- Works with GitHub Copilot
- Actively maintained — repository updated in the last 90 days
Author
Topics
Frequently asked questions
- What does agentic-eval do?
- Patterns and techniques for evaluating and improving AI agent outputs. Use this skill when: - Implementing self-critique and reflection loops - Building evaluator-optimizer pipelines for quality-critical generation - Creating test-driven code refinement workflows - Designing rubric-based or LLM-as-judge evaluation systems - Adding iterative improvement to agent outputs (code, reports, analysis) - Measuring and improving agent response quality
- How do I install agentic-eval?
- Run git clone https://github.com/github/awesome-copilot ~/.claude/skills/agentic-eval in your agent, then reload your skills. Review the source at https://github.com/github/awesome-copilot before installing.
- Is agentic-eval free to use?
- Yes. agentic-eval is free and open source under the MIT license, so you can read, run, and adapt it within that license's terms.
- Which agents does agentic-eval work with?
- agentic-eval lists GitHub Copilot — named in its SKILL.md, its plugin marketplace or its repository topics. Skills use the portable SKILL.md format, so other agents that read that format may load it too.
- Where does agentic-eval come from?
- agentic-eval ships inside github/awesome-copilot, a repository that contains 41 catalogued skills in total. The repository's 37,982 GitHub stars apply to that whole collection, not to this skill on its own.
