Set up your AI-powered Canvas assistant in just 5 minutes
You'll need an API token from your Canvas account to enable the integration.
Syllabus works with multiple AI platforms. Choose your preferred one:
npm install -g @anthropic/claude-codeclaude mcp add canvas -- npx canvas-mcp-toolWhen prompted, enter:
CANVAS_API_TOKEN: your_token_hereCANVAS_BASE_URL: https://canvas.instructure.comReplace with your institution's Canvas URL if different
claude --mcp canvasNavigate to:
~/Library/Application Support/Claude/claude_desktop_config.jsonOn Windows: %APPDATA%/Claude/
{
"mcpServers": {
"canvas": {
"command": "npx",
"args": ["canvas-mcp-tool"],
"env": {
"CANVAS_API_TOKEN": "your_token_here",
"CANVAS_BASE_URL": "https://canvas.instructure.com"
}
}
}
}Quit and reopen Claude Desktop to load the new configuration.
CANVAS_API_TOKEN=your_token CANVAS_BASE_URL=https://canvas.instructure.com npx canvas-mcp-toolUse this method for testing or integration with other MCP-compatible tools.
Once connected, try these natural language queries with your AI assistant:
| Variable | Description | Required |
|---|---|---|
| CANVAS_API_TOKEN | Your Canvas API access token | Yes |
| CANVAS_BASE_URL | Canvas instance URL | Yes |
| UBC_MODE | Enable UBC-specific features | No |
| DEBUG | Enable debug logging | No |
If you're getting authentication errors, ensure your token has the necessary permissions and hasn't expired.
Learn about Canvas API permissionsVerify your CANVAS_BASE_URL matches your institution's Canvas domain exactly, including https://.
If using Claude Desktop, ensure the configuration file is valid JSON and restart the app completely.