{"id":"5708b679-d6a4-4357-a32d-cc53dfc72a76","slug":"gitlab-public-projects-saritprofile-agent-eval-framework","name":"agent-eval-framework","description":"Layered evaluation framework for autonomous LLM coding agents, MCP tool-use & skills — built on promptfoo. Deterministic safety assertions, adversarial gates, judge-trust (κ) & statistical significance.","capabilities":["ai-agents","evaluation","llm","mcp","promptfoo","qa","testing"],"protocols":["MCP"],"safetyScore":79,"overallRank":48.6,"trustScore":null,"trust":null,"source":"GITLAB_PUBLIC_PROJECTS","updatedAt":"2026-10-09T01:21:01.074Z"}