Canvas Mcp

Overview

Canvas Mcp is a local-first Model Context Protocol server for Canvas LMS, using stdio transport with no third-party broker. It lets any MCP-compatible AI assistant call Canvas tools directly, with the model deciding when and how to use them in conversation or agent workflows.

The server runs locally via the Python package canvas-local-mcp, installed through uvx. You'll need to set CANVAS_BASE_URL and CANVAS_TOKEN as environment variables for API authentication. Version v0.1.3 is available now under the MIT license, with source code at admin978/canvas-mcp on GitHub.

Canvas Mcp is categorized as Design on mcp.site and integrates with MCP clients including Claude Desktop, Claude Code, Cursor, VS Code, Windsurf, and Zed. If you maintain this server, claim your listing to earn the Verified badge and ensure the information stays accurate.

Installation

// claude_desktop_config.json
{
  "mcpServers": {
    "canvas_mcp": {
      "command": "uvx",
      "args": [
        "canvas-local-mcp"
      ],
      "env": {
        "CANVAS_BASE_URL": "<canvas-base-url>",
        "CANVAS_TOKEN": "<canvas-token>"
      }
    }
  }
}

Requires CANVAS_BASE_URL, CANVAS_TOKEN