Skip to main content

OpenClaw

OpenClaw, running on a brain you own.

The open-source harness this product runs inside, with your memory attached.

OpenClaw is the open-source agent harness gbrain itself runs inside, so this is the pairing the whole thing was built around. Run it yourself and it starts with no memory of anything before the session you are in.

Point it at a gbrain workspace and it gets two things. Memory, as a place to write what it learns and read it back next week. And tools, over one MCP server rather than a folder of configs, each with its own key on disk.

You choose which tools it sees. The token in its config can be revoked without touching the harness.

One endpoint connects OpenClaw

Not one per service. What this client can call is decided by its permission, not by the address, and the token comes from the client's own page in the workspace, minted once and shown once.

Endpoint
https://gbrain.io/mcp
Header
Authorization: Bearer <token>

Every application OpenClaw gets

Connect one in the browser and it answers here, under the same permission and the same record as everywhere else.

The other places a connection answers are on Works with , and Tools explains the permission, the ceiling and the record they share.