Canvas MCP Server

MCP server for Canvas LMS with 88 tools and 8 agent skills. Works with Claude Desktop, Cursor, Codex, Windsurf, and 40+ other agents.

bashCopy code
npx skills add vishalsachdev/canvas-mcp

For AI Agents

Canvas MCP provides 88 tools for interacting with Canvas LMS. Tools are organized by user type:

Skills: canvas-course-qc (pre-semester audit), canvas-accessibility-auditor (WCAG compliance), canvas-course-builder (scaffold courses from specs/templates).

Decision tree: Simple query → MCP tools. Batch grading (10+) → bulk_grade_submissions. Complex bulk (30+) → execute_typescript.

Quick Reference

Course identifiers: Canvas ID (12345), course code (badm_350_120251_246794), or SIS ID

Cannot do: Create/delete courses, modify course settings, access other users' data, create/update rubrics (use Canvas UI)

Rate limits: ~700 requests/10 min. Use max_concurrent=5 for bulk operations.

Full documentation: AGENTS.md | tools/TOOL_MANIFEST.json | tools/README.md

Overview

The Canvas MCP Server bridges the gap between AI assistants and Canvas Learning Management System, providing both students and educators with an intelligent interface to their Canvas environment. Built on the Model Context Protocol (MCP), it enables natural language interactions with Canvas data through any MCP-compatible client.

Latest Release: v1.3.0

Released: May 2026 | Full Changelog | All Releases

• create_rubric — Programmatic rubric creation with criteria, ratings, and optional assignment association (PR #100)
• read_course_file — Read Canvas file content for remote MCP deployments (@DomBarker99, PR #90)
• Event-loop fix — Resolves "Event loop is closed" on user-scoped tools (get_my_todo_items, get_my_upcoming_assignments, etc.) (PR #99)
• ⚠️ Bulk-delete safety — bulk_delete_announcements now caps at 25 IDs by default; pass limit=N to override or dry_run=True to preview. Existing callers passing >25 IDs must add limit=N explicitly. (PR #96)
• Maintenance — Drop unused fastmcp dep, pin mcp>=1.26,<2, prune ~30 transitive deps (PR #93)

v1.2.0 — Role-Based Tool Filtering (@Promithius-DR, PR #84), Accessibility Remediation (fix_accessibility_issues, scanner expanded 4→20 checks), Security Hardening (path traversal/symlink protections), Windows Support for execute_typescript (PR #85), CI consolidation (11→8 checks)

v1.1.0 — Hosted Server (mcp.illinihunt.org), Learning Designer tools + 3 skills, Agent Skills on skills.sh, File Management (@Metzpapa, PR #75), Token Optimization, Generic Distribution

v1.0.8 — Security Hardening (PII sanitization, audit logging, sandbox-by-default), Ruff linting, 235+ tests

v1.0.7 — Assignment Update Tool (update_assignment), complete CRUD, 9 tests

v1.0.6 — Module Management (7 tools), Page Settings (2 tools), 235+ tests

v1.0.5 — Claude Code Skills, GitHub Pages site

v1.0.4 — Code Execution API (99.7% token savings), Bulk Operations, MCP 2.14 compliance

For Students 👨‍🎓

Get AI-powered assistance with:

• Tracking upcoming assignments and deadlines
• Monitoring your grades across all courses
• Managing peer review assignments
• Accessing course content and discussions
• Organizing your TODO list

→ Get Started as a Student

For Educators 👨‍🏫

Enhance your teaching with:

• Assignment and grading management
• Student analytics and performance tracking
• Discussion and peer review facilitation
• FERPA-compliant student data handling
• Bulk messaging and communication tools

→ Get Started as an Educator

For Learning Designers 🎨

AI-powered course design and quality assurance:

• Course scaffolding — Build entire course structures from specs, templates, or by cloning existing courses
• Quality audits — Pre-semester QC checks for structure, content, publishing, and completeness
• Accessibility compliance — 20-check WCAG scanner (headings, tables, scope, contrast, alt text, links, captions, DesignPLUS migration), prioritized reports, guided remediation, and verification
• Course structure analysis — Full module→items tree in a single call for rapid course review

3 dedicated skills (canvas-course-qc, canvas-accessibility-auditor, canvas-course-builder) plus the get_course_structure tool.

🤖 Agent Skills

Pre-built workflow recipes that teach AI agents how to use Canvas MCP tools effectively. Available for 40+ coding agents via skills.sh, or as Claude Code-specific slash commands.

Install via skills.sh (Any Agent)

bashCopy code
npx skills add vishalsachdev/canvas-mcp

This launches an interactive picker to install skills into your agent of choice (Claude Code, Cursor, Codex, OpenCode, Cline, Zed, and many more).

Install a specific skill:

bashCopy code
npx skills add vishalsachdev/canvas-mcp -s canvas-week-plan

Claude Code Slash Commands

If you use Claude Code, the same workflows are also available as slash commands:

Copy code
You: /canvas-morning-check CS 101
Claude: [Generates comprehensive course status report]

You: /canvas-week-plan
Claude: [Shows prioritized weekly assignment plan]

Claude Code skills are located in .claude/skills/ and can be customized for your workflow.

Want a custom skill? Submit a request describing your repetitive workflow!

🔒 Privacy & Data Protection

For Educators: FERPA Compliance

Complete FERPA compliance through systematic data anonymization when working with student data:

• Source-level data anonymization converts real names to consistent anonymous IDs (Student_xxxxxxxx)
• Automatic email masking and PII filtering from discussion posts and submissions
• Local-only processing with configurable privacy controls (ENABLE_DATA_ANONYMIZATION=true)
• FERPA-compliant analytics: Ask "Which students need support?" without exposing real identities
• De-anonymization mapping tool for faculty to correlate anonymous IDs with real students locally

All student data is anonymized before it reaches AI systems. See Educator Guide for configuration details.

For Students: Your Data Stays Private

• Your data only: Student tools access only your own Canvas data via Canvas API's "self" endpoints
• No credential storage: In hosted mode, your Canvas token is sent as an HTTP header per-request and never stored on the server. In local mode, everything runs on your machine.
• No tracking: Your Canvas usage and AI interactions remain private
• No anonymization needed: Since you're only accessing your own data, there are no privacy concerns

Use Without Installing (Hosted Server)

Connect to the hosted Canvas MCP server — no Python, no cloning, no setup. Just add a URL and your Canvas credentials to your MCP client.

All you need: A Canvas API token and your institution's Canvas URL.

Claude Desktop's claude_desktop_config.json expects stdio servers (command + args), so use an HTTP bridge:

jsonCopy code
{
  "mcpServers": {
    "canvas": {
      "command": "npx",
      "args": [
        "mcp-remote",
        "https://mcp.illinihunt.org/mcp",
        "--transport",
        "http-only",
        "--header",
        "X-Canvas-Token: ${CANVAS_TOKEN}",
        "--header",
        "X-Canvas-URL: ${CANVAS_URL}"
      ],
      "env": {
        "CANVAS_TOKEN": "your_canvas_api_token",
        "CANVAS_URL": "https://your-school.instructure.com/api/v1"
      }
    }
  }
}

Requires Node.js (npx).

Add to your MCP client configuration:

jsonCopy code
{
  "mcpServers": {
    "canvas": {
      "url": "https://mcp.illinihunt.org/mcp",
      "headers": {
        "X-Canvas-Token": "your_canvas_api_token",
        "X-Canvas-URL": "https://your-school.instructure.com/api/v1"
      }
    }
  }
}

bashCopy code
claude mcp add canvas \
  --transport http \
  --url https://mcp.illinihunt.org/mcp \
  --header "X-Canvas-Token: your_canvas_api_token" \
  --header "X-Canvas-URL: https://your-school.instructure.com/api/v1"

Get your Canvas API token: Canvas → Account → Settings → New Access Token

Find your Canvas URL: It's your institution's Canvas domain with /api/v1 appended (e.g., https://canvas.illinois.edu/api/v1). Check the URL bar when you log into Canvas.

Your credentials are sent as HTTP headers with each request — they are never stored on the server. All 88 tools work the same as local installation.

Privacy note: The hosted server does not store or log credentials or Canvas data. However, data passes through a third-party VPS in transit. Educators handling FERPA-protected student data should use the local installation instead. The hosted server is ideal for students (who only access their own data) and for trying out Canvas MCP before installing locally.

Prerequisites (Local Installation)

• Python 3.10+ - Required for modern features and type hints
• Canvas API Access - API token and institution URL
• MCP Client - Any MCP-compatible client (Claude Desktop, Cursor, Zed, Windsurf, Continue, etc.)

Supported MCP Clients

Works with any MCP-compatible client: Claude Desktop, Cursor, Zed, Windsurf, Continue, Replit, Copilot Studio, and more.

Canvas MCP is compliant with Canvas LMS API 2024-2026 requirements (User-Agent header, per_page pagination). Works with Canvas Cloud and self-hosted instances.

Local Installation

1. Install Dependencies

bashCopy code
# (Recommended) Use a dedicated virtualenv so the MCP binary is in a stable location
python3 -m venv .venv
. .venv/bin/activate

# Install the package editable
pip install -e .

2. Configure Environment

bashCopy code
# Copy environment template
cp env.template .env

# Edit with your Canvas credentials
# Required: CANVAS_API_TOKEN, CANVAS_API_URL

Get your Canvas API token from: Canvas → Account → Settings → New Access Token

Note for Students: Some educational institutions restrict API token creation for students. If you see an error like "There is a limit to the number of access tokens you can create" or cannot find the token creation option, contact your institution's Canvas administrator or IT support department to request API access or assistance in creating a token.

3. MCP Client Configuration

Canvas MCP works with any MCP-compatible client. Below are configuration examples for popular clients:

Configuration file location:

• macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
• Windows: %APPDATA%\Claude\claude_desktop_config.json

Configuration:

jsonCopy code
{
  "mcpServers": {
    "canvas-api": {
      "command": "/absolute/path/to/canvas-mcp/.venv/bin/canvas-mcp-server"
    }
  }
}

Note: Use the absolute path to your virtualenv binary to avoid issues with shell-specific PATH entries (e.g., pyenv shims).

Configuration file location:

• macOS/Linux: ~/.cursor/mcp_config.json
• Windows: %USERPROFILE%\.cursor\mcp_config.json

Configuration:

jsonCopy code
{
  "mcpServers": {
    "canvas-api": {
      "command": "/absolute/path/to/canvas-mcp/.venv/bin/canvas-mcp-server"
    }
  }
}

Configuration: Add to Zed's settings.json (accessible via Settings menu)

jsonCopy code
{
  "context_servers": {
    "canvas-api": {
      "command": {
        "path": "/absolute/path/to/canvas-mcp/.venv/bin/canvas-mcp-server",
        "args": []
      }
    }
  }
}

Configuration file location:

• macOS: ~/Library/Application Support/Windsurf/mcp_config.json
• Windows: %APPDATA%\Windsurf\mcp_config.json

Configuration:

jsonCopy code
{
  "mcpServers": {
    "canvas-api": {
      "command": "/absolute/path/to/canvas-mcp/.venv/bin/canvas-mcp-server"
    }
  }
}

Configuration: Add to Continue's config.json (accessible via Continue settings)

jsonCopy code
{
  "mcpServers": {
    "canvas-api": {
      "command": "/absolute/path/to/canvas-mcp/.venv/bin/canvas-mcp-server"
    }
  }
}

For other MCP-compatible clients, the general pattern is:

1. Locate your client's MCP configuration file
2. Add a server entry with:
   • Server name: canvas-api (or any name you prefer)
   • Command: Full path to canvas-mcp-server binary
   • Optional args: Additional arguments if needed

Consult your client's MCP documentation for specific configuration format and file locations.

Windows users: Replace forward slashes with backslashes in paths (e.g., C:\Users\YourName\canvas-mcp\.venv\Scripts\canvas-mcp-server.exe)

Verification

Test your setup:

bashCopy code
# Test Canvas API connection
canvas-mcp-server --test

# View configuration
canvas-mcp-server --config

# Start server (for manual testing)
canvas-mcp-server

Available Tools

The Canvas MCP Server provides a comprehensive set of tools for interacting with the Canvas LMS API. These tools are organized into logical categories for better discoverability and maintainability.

Tool Categories

Student Tools (New!)

• Personal assignment tracking and deadline management
• Grade monitoring across all courses
• TODO list and peer review management
• Submission status tracking

Shared Tools (Both Students & Educators)

1. Course Tools - List and manage courses, get detailed information, generate summaries with syllabus content
2. Discussion & Announcement Tools - Manage discussions, announcements, and replies
3. Page & Content Tools - Access pages, modules, and course content

Educator Tools 4. Assignment Tools - Handle assignments, submissions, and peer reviews with analytics 5. Rubric Tools - List rubrics, associate with assignments, and grade submissions (including bulk_grade_submissions for efficient batch grading). Note: Create/update rubrics via Canvas web UI due to API limitations. 6. User & Enrollment Tools - Manage enrollments, users, and groups 7. Analytics Tools - View student analytics, assignment statistics, and progress tracking 8. Messaging Tools - Send messages and announcements to students

Developer Tools 9. Discovery Tools - Search and explore available code execution API operations with search_canvas_tools and list_code_api_modules 10. Code Execution Tools - Execute TypeScript code with execute_typescript for token-efficient bulk operations (99.7% token savings!)

📖 View Full Tool Documentation for detailed information about all available tools.

Code Execution API

For bulk operations (30+ items), Canvas MCP supports TypeScript code execution with 99.7% token savings compared to traditional tool calling.

Use search_canvas_tools to discover available operations, then execute_typescript to run them locally. Code runs in a secure sandbox by default (network blocked, env filtered, resource limits). Works on macOS, Linux, and Windows.

Bulk Grading Example

typescriptCopy code
import { bulkGrade } from './canvas/grading/bulkGrade';

await bulkGrade({
  courseIdentifier: "60366",
  assignmentId: "123",
  gradingFunction: (submission) => {
    const notebook = submission.attachments?.find(f =>
      f.filename.endsWith('.ipynb')
    );
    if (!notebook) return null;
    return { points: 100, comment: "Great work!" };
  }
});

Security Modes

See Bulk Grading Example for a detailed walkthrough.

Usage

MCP clients start the server automatically. Just ask naturally:

• "What's due this week?" / "Show my grades" / "What peer reviews do I need?"
• "Who hasn't submitted Assignment 3?" / "Send reminders to missing students"

Quick start guides: Student | Educator | Real-World Workflows | Troubleshooting

Documentation

• Tool Documentation — Complete reference for all 88 tools
• Student Guide — Getting started as a student
• Educator Guide — FERPA compliance and educator workflows
• Bulk Grading Example — Token-efficient batch grading walkthrough
• Development Guide — Architecture and contributing

Built on FastMCP with async httpx, pydantic validation, and python-dotenv configuration. Modern src/ layout with pyproject.toml. Full type hints, connection pooling, smart pagination, and rate limiting. 328 tests. ruff + black for code quality.

Troubleshooting

If you encounter issues:

1. Server Won't Start - Verify your Configuration setup: .env file, virtual environment path, and dependencies
2. Authentication Errors - Check your Canvas API token validity and permissions
3. Connection Issues - Verify Canvas API URL correctness and network access
4. Debugging - Check your MCP client's console logs (e.g., Claude Desktop's developer console) or run server manually for error output

Security

Four layers of runtime security, all enabled by default:

FERPA-compliant anonymization for educators: ENABLE_DATA_ANONYMIZATION=true. See Educator Guide for details.

Publishing

Published to PyPI, MCP Registry, and skills.sh (agent skills). Releases are automated via GitHub Actions — tag a version (git tag vX.Y.Z && git push origin vX.Y.Z) and CI handles the rest.

Contributing

Contributions are welcome! Feel free to:

• Submit issues for bugs or feature requests
• Create pull requests with improvements
• Share your use cases and feedback

Contributors

Thanks to everyone who has contributed to Canvas MCP:

• @DomBarker99 — read_course_file tool for remote MCP deployments (#90)
• @Promithius-DR — Role-based tool filtering and tool annotations (#84)
• @Metzpapa — File download and listing tools (#75)
• @JCSnap — Student tool bug fixes (#72, #73)

License

This project is licensed under the MIT License - see the LICENSE file for details.

Created by Vishal Sachdev