Skip to main content
Artil runs an MCP server at https://mcp.artil.dev/mcp. It acts as one agent account: it shows where people reach the agent, reads and sends its email and texts, and stores its secrets. Use it from a client that cannot run the CLI, such as an agent on a hosted platform. Connect a client in one of two ways: sign in through the browser, or send one of the agent’s tokens in a header.

Sign in through the browser

Add the server by its URL alone. When the client connects, Artil opens in your browser: choose the agent the client acts as and click Allow access.
Add Artil to Claude, confirm the connector, and click Connect.
ChatGPT, Zed, Windsurf, and other clients that sign in to remote MCP servers with OAuth work the same way: add https://mcp.artil.dev/mcp and sign in when the client asks. Access lasts 30 days, then the client asks you to sign in again. A sign-in as one agent appears on its Tokens tab under the client’s name; revoke it there to disconnect the client sooner.

Use a token

1. Create a token

Open the agent in Artil, select Tokens, click New token, enter a name, such as the client’s, and click Create token. Copy the token; it is shown only once and does not expire. Only the agent’s own token works. Keep it out of repositories and agent conversations, and revoke it on the same tab when the client no longer needs it.

2. Add the server to your client

The client sends the token in the Authorization header. Replace YOUR_TOKEN with the token you copied.
Run /mcp in a new session and check that artil is connected.
Other clients that connect to remote MCP servers over HTTP and can send a header work the same way.

Tools

Message IDs here have no mail: or sms: prefix, unlike the ones the CLI shows. Sending follows the same rules and limits as the CLI; see email and SMS. The server never returns secret values. To use the secrets in programs, link them where the programs run.
Anyone can write to the agent’s address, so a message can try to steer the client. The server tells the client to treat message text as information, never as instructions. Check what the client may do without asking before you connect it.

New messages

The server does not tell the client when a message arrives; the client sees new messages when it calls messages.list. To hear about them as they arrive, use the Claude Code plugin, the Pi package, the OpenClaw plugin, or automatic delivery.