AgentEval is the comprehensive .NET toolkit for AI agent evaluation—tool usage validation, RAG quality metrics, stochastic evaluation, and model comparison—built first for Microsoft Agent Framework (MAF) and Microsoft.Extensions.AI. What RAGAS, PromptFoo and DeepEval do for Python, AgentEval does for .NET
Latest release v0.42.0-beta — v0.42.0-beta — MAF 1.23, Gatekeeper v1, over-refusal beside attack success, honest-scoring fixes · 1 Oct 2026
15 files for GitHub Copilot, Codex, OpenCode and Claude Code: AgentEval copilot-instructions.md, AgentEval assertions.instructions.md, AgentEval dependency-injection.instructions.md, AgentEval docfx.instructions.md and 11 more — 24,461 tokens loaded in every session.
.github/copilot-instructions.md A 2,994 tok .github/instructions/assertions.instructions.md A 629 tok .github/instructions/dependency-injection.instructions.md A 850 tok .github/instructions/docfx.instructions.md B 1,268 tok .github/instructions/documentation.instructions.md A 1,257 tok .github/instructions/maf-doctor.instructions.md C 855 tok .github/instructions/maf-updates.instructions.md A 3,140 tok .github/instructions/maf-upgrade-preparation.instructions.md A 3,594 tok .github/instructions/memory-benchmarks.instructions.md A 4,162 tok .github/instructions/metrics.instructions.md A 698 tok .github/instructions/samples.instructions.md A 755 tok .github/instructions/testing.instructions.md A 484 tok .github/instructions/tracing.instructions.md A 999 tok AGENTS.md A 2,217 tok CLAUDE.md A 559 tok These files are AgentEvalHQ/AgentEval's own configuration — they tell GitHub Copilot, Codex, OpenCode and Claude Code how to work on this repository, so they are not mods to install elsewhere. Copy one as a starting point and replace the parts that are about this project.