canvas-mcp

canvas-mcp addresses the gap between AI agents and Canvas LMS. Students and educators usually perform course management, assignment submission, and grade checking manually in Canvas. With canvas-mcp, AI agents can automate these interactions.

The project exposes 90 MCP tools organized by user role (student, instructor, etc.), covering course info, assignment management, grade retrieval, and other typical LMS workflows. Install with `npx skills add vishalsachdev/canvas-mcp`. It works with Claude Desktop, Cursor, Codex, Windsurf, and 40+ other agent clients.

Strengths include broad tool coverage and role-based organization, plus easy installation. The main dependency is an existing Canvas LMS instance with valid API credentials; you need a Canvas account and token to use it. Make sure your Canvas environment is ready before adopting.

Best suited for students and educators already on Canvas LMS who want to automate LMS interactions through AI agents; not suitable for users without Canvas access.