The repository publishes a demo token and demo API URL, so the whole evaluation is build the binary, paste two values, restart the client — nothing to sign up for. Past that point it is a front end to your Metoro account, so its usefulness depends entirely on having the eBPF agents running against the cluster you care about.
A Go MCP server that exposes Metoro's observability APIs to an AI client. Metoro instruments microservices in Kubernetes with eBPF agents, so the telemetry exists without changing any application code — this server puts that data in the conversation.
- Questions about a live Kubernetes cluster answered from the same APIs the Metoro frontend uses
- Deep telemetry from eBPF instrumentation, without adding libraries or redeploying services
- A public demo cluster you can point it at before committing to an account
Go installed, so you can clone the repository and build the binary. Then set `METORO_AUTH_TOKEN` from your Metoro user settings and `METORO_API_URL` for your region in the client config. To try it without an account, point both at the demo cluster using the demo token published in the repository.
Build from source — clone the repository and build it, then point your client at the binary