Agent Explorer: test integrations with a real AI agent

Agent Explorer is a new way to test integrations alongside the existing Explorer. Where Explorer lets you browse an integration and call tools yourself, Agent Explorer hands the integration to a real AI model and shows you what it actually does with it. You can run the same integration across Claude, ChatGPT, Gemini, Grok, and others to see how each model behaves before anything goes to production. Every test run goes through the same access control and tracing as live traffic.

Ready to build with Metorial?

Connect any AI agent to any tool or data source. Govern every action.