Give your AI agent the keys.

Browser-based builders (Lovable, Bolt, v0)

AI app builders that run in the browser connect to a remote MCP server by URL. Point yours at the Canner endpoint below and authenticate with a Canner API token, then ask your agent to “save my app to Canner” — it can create a project and deploy your code straight from the chat, onto Canadian infrastructure. In Lovable: Connectors → Chat connectors → New MCP server. In Bolt and v0: add a custom MCP server in settings.

In your dashboard's API tokens, choose the “MCP connector (deploy only)” type — it's scoped to project + deploy access with no billing and no token management, so a credential you paste into a third-party builder can't do more than deploy.

Local clients (Claude Desktop, Cursor, Claude Code)

Desktop agents run the server over npx — nothing to install globally. Add it to your MCP client's config with a Canner API token. In Claude Desktop, edit claude_desktop_config.json; in Cursor, edit .cursor/mcp.json. Then restart the client and the Canner tools appear.

{
  "mcpServers": {
    "canner": {
      "command": "npx",
      "args": ["-y", "@canner-ca/mcp"],
      "env": { "CANNER_TOKEN": "cnr_your_token_here" }
    }
  }
}

Already used canner login with the CLI? You can drop the CANNER_TOKEN line — the server reads the token the CLI stored.

Tools

Thirteen tools, each wrapping a documented REST endpoint — including canner_create_project and canner_save_app, which deploy your code from the agent:

Zero dependencies

The server speaks the MCP stdio transport directly — no third-party packages. That's the same supply-chain discipline as the rest of Canner: the code you run inside your agent ships nothing extra to audit, and never reaches outside Canada in normal operation.

Safe by default

Destructive operations — deleting a project, revoking a token, dropping a database or domain — are deliberately not exposed, so an agent can't take them by mistake. Do those from the dashboard or the CLI.

What it stores and sends

The server writes nothing to your machine. It reads a token from the CANNER_TOKEN environment variable, or — if you've already run canner login — from the credentials file the CLI wrote at ~/.canner/credentials. Requests go only to api.canner.ca. There is no telemetry, analytics, or crash reporting. Run via npx, nothing is installed globally; to remove it, delete the entry from your MCP client's config file.

Use of the MCP server and the Canner service is governed by our Terms of Service and Privacy Policy.