The same tools, in any assistant that speaks MCP
Claude Code, Cursor and Codex speak MCP. Point one here and it reaches the same accounts.
gbrain.io
Claude Code, Cursor and Codex speak MCP. Point one at this endpoint and it reaches the same accounts, under the same permission and daily limit, writing to the same activity list.
claude mcp add gbrain https://gbrain.io/mcp --header "Authorization: Bearer <token>"
The token comes from the connection's own page in the workspace, and is shown once.
An assistant is handed only the tools its permission allows. Anything you didn't grant isn't listed, offered, or callable.
Refusals come back as structured data rather than prose, so an assistant reads what was missing and asks for it instead of guessing at a retry.
Calls land on the same activity list as the terminal's, against the same daily limit.
