Canvas MCP

🚀 Canvas Canvas MCP Server Assistant

This Model Context Protocol (MCP) server allows you to interact with Canvas/Instructure courses and assignments without leaving your LLM (e.g., Claude Desktop).

It enables you to programmatically retrieve, search, and summarize course and assignment information. For example, you can check the due dates of upcoming assignments:

✨ Features

Tools

1. List Courses

   • Retrieve a list of courses.
   • Filter by course status (active, completed, or all).

2. Search Assignments

   • Search for assignments across all courses.
   • Filters include:
     • Search query.
     • Due date range.
     • Specific courses.
     • Include/exclude completed courses.

3. Get Assignment Details

   • Get detailed information for a specific assignment.
   • Multiple output formats (full HTML, plain text, Markdown).

Resources

• Assignment Content: Retrieve full assignment details in Uniform Resource Identifier format.

📋 Prerequisites

• Node.js
• Canvas LMS account
• Canvas API token
• Canvas base URL

⚙️ Environment Setup

Set the following environment variables:

• CANVAS_API_TOKEN: Your Canvas API access token (see instructions below).
• CANVAS_DOMAIN: The domain name of your Canvas institution (e.g., canvas.youruniversity.edu).

How to Get Your Canvas API Token

1. Log in to Canvas.
2. Go to Account > Settings.
3. Scroll to the "Approved Integrations" section.
4. Click "New Access Token".
5. Copy the generated token.

For more details, refer to these instructions from Canvas.

⚠️ Security Note

Keep your CANVAS_API_TOKEN confidential. Do not commit it to version control.

📦 Installation

1. Clone the repository and install dependencies

   Clone the repository:

   git clone https://github.com/mbcrosier/canvas-mcp.git
   cd canvas-mcp
   

   Install dependencies. If you encounter errors, make sure you have Node.js installed.

   npm install
   

2. Connect to the MCP server

   Copy the following JSON and replace the appropriate {{PATH}} values:

   {
       "mcpServers": {
           "canvas": {
               "command": "node",
               "args": ["/absolute/path/to/canvas-mcp/src/index.ts"], // Navigate to the src repository, run `pwd`, and enter the output here
               "env": {
                   "CANVAS_API_TOKEN": "your_api_token_here",
                   "CANVAS_DOMAIN": "myschool.instructure.com"
               }
           }
       }
   }
   

   For Claude, save this file as claude_desktop_config.json in your Claude Desktop configuration directory:

   ~/Library/Application Support/Claude/claude_desktop_config.json
   

   For Cursor, save this file as mcp.json in your Cursor configuration directory:

   ~/.cursor/mcp.json
   

3. Restart Claude Desktop / Cursor

   Open Claude Desktop, and you should now see Canvas as an available integration.

🛠️ MCP Tools

• canvas: Connect to the Canvas platform and perform various operations, such as retrieving course and assignment information, submitting assignments, etc.

🚧 Troubleshooting

If you encounter issues, check the following:

• Ensure that Node.js and dependencies are installed correctly.
• Ensure that the API token and domain name are set correctly.
• Check the console output for more error details.

📄 Disclaimer

This project aims to provide a basic tool for interacting with the Canvas platform. Please read the relevant documentation carefully and ensure compliance with Canvas's terms of service and privacy policy before use.