AI agents are powerful, but testing them is guesswork. Agentest brings structure with scenario-based tests, simulated users, tool-call mocks, and LLM-as-judge evaluation. Think Vitest or Playwright for agent workflows. Run tests locally or in CI and ship reliable agents with confidence.
No reviews yetBe the first to leave a review for Agentesting
Maker
📌
Hey everyone! 👋
I’m Raffael, a developer who’s been diving into AI agents over the past year.
While building agents, I kept running into the same problem: testing them felt like guesswork. Tools like LangSmith and Langfuse are great for evaluation, but I was missing proper testing workflows, and manual testing didn’t scale. So I built Agentest.
It’s still very early, and I’d really love your feedback. If this resonates, a ⭐ on GitHub or even a PR would mean a lot 🙌
Also super curious, how are you currently testing your agents?