Canvas MCP Server

# Educator Quick Start Guide This guide shows the most common tasks educators use Canvas MCP for. ## Setup 1. Install Canvas MCP following the [README](../README.md#installation) 2. Configure your `.env` file with Canvas API token 3. **For FERPA compliance**, add: ``` ENABLE_DATA_ANONYMIZATION=true ``` 4. Restart Claude Desktop ## Common Tasks ### Check Submission Status ``` Which students haven't submitted Assignment 5 in CS 101? ``` Claude will show you who's missing submissions and help you follow up. ### Grade Assignments with Rubrics ``` I need to grade Assignment 3 in BADM 350 using the rubric ``` Claude can: - Show you the rubric criteria - Help you grade submissions - Apply consistent grading across students ### Send Reminders ``` Send a reminder to students who haven't submitted the midterm project ``` Claude will: - Identify students with missing submissions - Draft a message - Send it through Canvas ### Analyze Participation ``` Who hasn't participated in the Discussion 4 in my Intro to Python course? ``` Get quick insights into discussion participation. ### Bulk Operations For large classes, use the code execution API: ``` Grade all 90 Jupyter notebook submissions for Assignment 2 by checking if they run without errors ``` This uses the bulk grading feature that's **99.7% more efficient** than traditional methods. ### Check Assignment Statistics ``` Show me statistics for Quiz 3 in ECON 102 ``` Get: - Average score - High/low scores - Submission rate - Common issues ## Advanced Workflows ### Weekly Check-in ``` For CS 225, show me: 1. Who's falling behind (missing multiple assignments) 2. Discussion participation rates 3. Average grades on the latest assignment ``` ### Grade Discussion with Rubric ``` Grade Discussion Topic 5 in BADM 350. Give 10 points for initial post, 5 points per peer review (must do 2), max 10 points for peer reviews. ``` Uses bulk discussion grading - super fast! ### Peer Review Analysis ``` Show me peer review completion for Assignment 4. Who hasn't done their reviews? ``` ### Course Health Check ``` Give me an overview of MATH 221: - How many students are current on assignments? - What's the average grade? - Who might need extra support? ``` ## FERPA Compliance Tips With `ENABLE_DATA_ANONYMIZATION=true`: - Student names appear as "Student_xxxxxxxx" - You can still identify patterns and get insights - Real identities stay private from AI processing - You keep a local mapping to correlate IDs with real students Example anonymized output: ``` Missing submissions: - Student_a8f7e23d (submitted 2/5 assignments) - Student_c9b21f84 (submitted 1/5 assignments) ``` ## Performance Tips **For small operations (< 10 students):** Just ask naturally **For bulk operations (> 30 students):** Mention you want to use code execution: ``` Use the bulk grading code API to grade all submissions for Assignment 6 ``` ## Common Pitfalls ❌ **Don't**: "Grade all assignments" (too vague) ✅ **Do**: "Grade Assignment 5 in CS 101 using the provided rubric" ❌ **Don't**: Send messages without reviewing ✅ **Do**: "Draft a message to students missing Assignment 3, let me review before sending" ❌ **Don't**: Assume old data is current ✅ **Do**: Canvas MCP fetches fresh data, but cache may be a few minutes old ## Troubleshooting **"Can't access student data"** - Check that your Canvas API token has the right permissions - Some institutions restrict educator access - contact your Canvas admin **"Bulk operations are slow"** - Use the code execution API for large batches - Mention "use bulk grading" in your request to Claude **"Anonymization isn't working"** - Verify `ENABLE_DATA_ANONYMIZATION=true` in your `.env` - Restart Claude Desktop after changing settings Need help? Check [EDUCATOR_GUIDE.md](../docs/EDUCATOR_GUIDE.md) or [open an issue](https://github.com/vishalsachdev/canvas-mcp/issues).