What is Canvas Mcp

Canvas-MCP is a server designed to help AI assistants like Claude access and manage content from the Canvas Learning Management System (LMS). It allows students to easily inquire about course-related information such as assignments and deadlines.

Use cases

Use cases for Canvas-MCP include enabling students to quickly access course information, asking questions about due assignments, and retrieving lecture notes, thereby enhancing the learning experience through AI assistance.

How to use

To use Canvas-MCP, you need to install the UV package manager, clone the repository, set up your environment, and configure your Claude Desktop to integrate with Canvas-MCP. Detailed setup instructions are provided in the README.

Key features

Key features of Canvas-MCP include automatic synchronization of courses and assignments, flexible document parsing from various formats (PDFs, Word docs, HTML), term filtering for current courses, and the option to exclude specific courses from indexing.

Where to use

Canvas-MCP is primarily used in educational settings where Canvas LMS is implemented, making it suitable for schools, colleges, and universities that utilize Canvas for course management.

Canvas AI Assistant with MCP

Overview

This repository contains our project for building a Model Context Protocol (MCP) server that integrates with the Canvas learning management system (LMS). The goal is to provide an AI-driven assistant (e.g., Claude) with structured access to course information—assignments, due dates, lecture notes—so that it can help students stay organized and succeed academically.

Getting Started

Mac:

0. Install uv if you haven’t already:

# if you're paranoid about piping to sh, go read the script by copy pasting the URL
curl -LsSf https://astral.sh/uv/install.sh | sh

1. Clone the repository:

git clone https://github.com/AdityaPrakash-26/Canvas-MCP.git && cd Canvas-MCP

2. Create the virtual environment:

uv venv --seed

3. Activate the virtual environment:

source .venv/bin/activate

4. Install the dependencies:

uv sync

Installing to Claude Desktop

If you don’t have Claude Desktop MCP setup, you can follow the Quickstart Guide to install it.

To install this MCP server in Claude Desktop:

1. Open Claude Desktop
2. Go to Settings > Advanced > Edit Configuration
3. (MAC ONLY) Add the following to your claude_desktop_config.json file in the tools section:
   (it is a key under mcpServers)
4. (WINDOWS ONLY) See instructions below

REPLACE $DIR with the absolute path to the directory where you cloned this repo. (MANDATORY!!!)
REPLACE $DIR_uv with the uv path. (MANDATORY!!!) you can find it by running

which uv

Windows

We recommend using WSL (Windows Subsystem for Linux) to use this software. Installation instructions can be founder here.

• Once installed, set up a new user in /home/<username>
• Once you have the virtual environment setup, modify your claude desktop app config to include the following:

Replace <USER> with your username.

Running

To run the mcp command with uv:

uv run mcp

Testing

Alternatively, you can test it with the MCP Inspector:

mcp dev src/canvas_mcp/server.py

Secrets

Clone .env.example to .env and add your Canvas API key.
Also change the variables in the .env file to match your Canvas instance.
“CANVAS_API_KEY”
“CANVAS_API_URL”

Repository & Kanban

• This public GitHub repository contains our code, documentation, and issues/tasks.
• We maintain a Kanban board under the “Projects” tab with tasks from our roadmap for the first sprint.

Link to the Repo: (https://github.com/AdityaPrakash-26/Canvas-MCP)

Link to the Kanban: (https://github.com/users/AdityaPrakash-26/projects/1)

License