Skip to content

Add agentic eval harness - #67

Merged
jlitola merged 5 commits into
mainfrom
opposite-hamburger
May 7, 2026
Merged

Add agentic eval harness#67
jlitola merged 5 commits into
mainfrom
opposite-hamburger

Conversation

@jlitola

@jlitola jlitola commented May 7, 2026

Copy link
Copy Markdown
Contributor

Summary

  • Add a human/agent-driven eval harness for running Claude and Codex against local or published GitHits MCP servers.
  • Add targeted workloads by tool family plus normalized tool-calls.json artifacts for reviewing actual tool use.
  • Document when to run agentic evals and clarify they are qualitative reviews, not CI gates.

Verification

  • bun test scripts/agent-eval.test.ts
  • bun test
  • bun run build
  • Live local Claude and Codex evals across the workload suite

@jlitola
jlitola merged commit da25feb into main May 7, 2026
3 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant