MCP

The Tilebox MCP server connects AI tools to Tilebox through the Model Context Protocol. It exposes tools for working with Tilebox datasets and workflows, and it includes documentation search so agents can retrieve current Tilebox context before answering questions or taking action.

Use MCP when your agent cannot easily use terminal commands. For coding agents with shell access, the CLI and Skills are the default setup. MCP is a good fit for web-based agents, chat tools, or hosted AI clients that support MCP but do not run local commands.

The CLI can build local Python workflow code, upload its artifact, and publish a new release. MCP cannot perform those steps, but it can deploy or undeploy an already-published release. Use the CLI and skills for the full code-to-runner development loop. With either interface, a compatible runner must be running on the target cluster for tasks to execute. Deploying a release does not start compute.

What the MCP server provides

The MCP server at https://mcp.tilebox.com/mcp provides tools for:

• Accessing and interacting with Tilebox datasets and workflows
• Searching the Tilebox documentation

When an MCP-capable client has the server configured, it can invoke relevant tools during a conversation. For example, when you ask about a dataset, the agent can query the Tilebox MCP server for the current schema and include that information in its response.

Authentication

The Tilebox MCP server uses OAuth. When your AI client connects, you are redirected to sign in with your Tilebox account and select a team. You do not need to configure API keys or custom headers for MCP authentication.

Configure an MCP client

Use https://mcp.tilebox.com/mcp with HTTP transport. Setup commands and configuration formats differ by client.

• Claude Code: Add the server, then run /mcp in Claude Code to authenticate:

  bash

  claude mcp add \
    --transport http "Tilebox" https://mcp.tilebox.com/mcp

• Amp: Add a personal remote definition to make the connection available across Amp clients, including orbs:

  bash

  amp mcp remote \
    --personal add Tilebox https://mcp.tilebox.com/mcp
  amp mcp remote --personal login Tilebox

  Open the sign-in URL printed by the login command. Alternatively, add the server in Amp’s MCP settings.

• Codex: Add the HTTP server and start the OAuth flow:

  bash

  codex mcp add tilebox \
    --url https://mcp.tilebox.com/mcp
  codex mcp login tilebox

• OpenCode: Run opencode mcp add, choose a remote server, name it tilebox, and enter https://mcp.tilebox.com/mcp. Then sign in:

  bash

  opencode mcp auth tilebox

• Cursor: Add this configuration to ~/.cursor/mcp.json, or merge the entry into its existing mcpServers object. Enable the server in Cursor and complete the OAuth sign-in flow.

  json

  {
    "mcpServers": {
      "tilebox": {
        "url": "https://mcp.tilebox.com/mcp"
      }
    }
  }

• Pi: Add the HTTP server, then sign in with OAuth:

  bash

  pi mcp add tilebox \
    --url https://mcp.tilebox.com/mcp
  pi mcp login tilebox

  Complete sign-in in your browser. If Pi is already running, use /reload to load the new server configuration.

For other clients, use their remote HTTP server setup with the same endpoint. Do not assume they accept Cursor’s JSON format. See the client documentation for Claude Code, Amp, Codex, OpenCode, Cursor, and Pi.

Use multiple teams

If you need to access multiple Tilebox teams, add a separate MCP server entry for each team. Each entry goes through its own OAuth flow and can be authorized for a different team. For example, in Cursor:

{
  "mcpServers": {
    "tilebox-team1": {
      "url": "https://mcp.tilebox.com/mcp"
    },
    "tilebox-team2": {
      "url": "https://mcp.tilebox.com/mcp"
    }
  }
}