Getting started with Statio

Sign up at app.statio.dev. The Free plan needs no card. The web dashboard is where you add servers and credentials, grant access, invite your team and read Activity.

Every person who uses an AI client with Statio installs the desktop app for macOS, Windows or Linux and signs in with their account. It connects that machine to the gateway, keeps its device key in the system keychain where one is available, writes the configuration for each AI client, and includes the mcphub CLI. On a Mac you can install it with Homebrew:

Add the API keys your agents call through, whether that is Stripe, GitHub, Slack or something internal. They are encrypted with AES-256-GCM at rest, fetched at call time inside the gateway, and never handed to an agent.

Turn individual tools on and off rather than whole servers. An agent allowed to read a database is not thereby allowed to delete from it, and the gateway checks that on every request before it decrypts anything.

In the desktop app, configure each client from Config in one click, or from the terminal with the mcphub CLI that comes with it. Either way it writes the correct MCP configuration for Claude Code, Claude Desktop, Cursor, Windsurf, VS Code, Cline, Continue, Zed and Amazon Q, so nobody hand-edits JSON. The web dashboard does not configure clients.

To install the CLI on its own, on macOS or Linux:

Agents sign in as themselves rather than borrowing a person’s token. Agent tokens last five minutes and the agent signs in again with its own credential, so a leaked token is soon useless, and you can suspend or revoke the agent from the dashboard within seconds.

Every tool call is logged in Activity with the identity behind it, a timestamp, the latency and the outcome. Audit a machine for MCP servers Statio does not manage, enroll devices to see them across the team in Fleet (Team plan and up), and stream the whole record to your SIEM (Business plan and up).