samikuikka/apo
Capability testing for agent harnesses: turn expert knowledge into executable specifications, run the real agent, get a binary pass/fail verdict with tests, traces, and deliverables.
GitHub repository with 5 stars and 3 forks.
Language: TypeScript
Topics: agent-evaluation, agent-evaluations, agent-testing, ai-agents, ai-sdk, developer-tools, e2e-testing, fastapi, llm, llm-evaluation