Skip to main content
Glama

Canvas MCP Server V2.0

# Canvas MCP Server v2.2.2 - Release Commands ## Quick Deploy (Automated) ```bash cd /Users/davidmontgomery/mcp-canvas-lms chmod +x deploy-fix.sh ./deploy-fix.sh ``` ## Manual Step-by-Step (if you prefer control) ### 1. Build the project ```bash cd /Users/davidmontgomery/mcp-canvas-lms npm run clean npm run build ``` ### 2. Git operations ```bash # Add all changes git add . # Check what's being committed git status # Commit with descriptive message git commit -m "fix: resolve console.log stdout pollution causing JSON parsing errors - Changed console.log to console.error in src/client.ts (request logging, error responses, retries) - Fixed tool execution logging in src/index.ts to use stderr - Eliminates 'Unexpected token C, [Canvas API' JSON parsing errors - MCP protocol now receives clean JSON communication over stdio - Version bump to 2.2.2 Fixes JSON parsing error spam in Claude Desktop" # Force push to main (be careful!) git push origin main --force ``` ### 3. Publish to npm ```bash # Publish the new version npm publish ``` ## What this fixes: ✅ Eliminates the constant "Unexpected token 'C', '[Canvas API'..." errors ✅ Stops the 10 errors/second JSON parsing spam in Claude Desktop ✅ Makes MCP communication clean (JSON only on stdout, logs on stderr) ✅ No breaking changes - purely a protocol fix ## Version Changes: - 2.2.1 → 2.2.2 (patch version for the console.log fix) - Updated in both package.json and src/index.ts - Added comprehensive CHANGELOG entry The core issue was debug logs polluting stdout which broke MCP's JSON communication protocol!

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/DMontgomery40/mcp-canvas-lms'

If you have feedback or need assistance with the MCP directory API, please join our Discord server