Study Cue
MCP Server guide

Use Claude with your study data.

The Study Cue MCP server connects Claude Desktop directly to your plans, resources, PDFs, and stats — so you can ask Claude anything about your studies without copy-pasting data.

Setup
01
Install the MCP package

Inside the backend2 directory, run: uv sync (or pip install mcp>=1.9.0). This adds the Anthropic MCP SDK that the server depends on.

cd backend2 && uv sync
02
Set your environment variables

The MCP server needs two env vars — your MongoDB connection string and (optionally) your Groq API key for any AI-powered tools you add later.

MONGODB_URI=mongodb://localhost:27017/studycue
GROQ_API_KEY=your-key-here
03
Add to Claude Desktop config

Open your Claude Desktop config file and add the study-cue block inside mcpServers. Update the cwd path to match where you cloned the project.

{
  "mcpServers": {
    "study-cue": {
      "command": "python",
      "args": ["-m", "backend2.mcp.run"],
      "cwd": "/path/to/study-cue",
      "env": {
        "MONGODB_URI": "mongodb://localhost:27017/studycue"
      }
    }
  }
}
04
Restart Claude Desktop

Fully quit and reopen Claude Desktop. The Study Cue tools appear in the tools panel — look for the hammer icon in any conversation.

05
Ask Claude about your study data

Claude can now read your plans, resources, PDF history, and stats. Just ask naturally — Claude picks the right tool automatically.

Available tools
get_active_study_planStudy Plan

Returns your active plan — subject, exam date, weekly tasks, and AI recommendations.

Try: "What is on my study plan for this week?"
list_resource_foldersResource Folders

Lists all your saved topic folders with resource counts.

Try: "What topics do I have resources saved for?"
get_resource_folderTopic Resources

Returns every saved resource (title, URL, type, description) for a specific topic.

Try: "Show me my saved resources for Biology."
list_pdfsPDF Documents

Lists all uploaded PDFs with page count and number of chat messages.

Try: "What PDFs have I uploaded to Study Cue?"
get_pdf_chat_historyPDF Chat History

Returns recent Q&A messages for a specific PDF, including the pages cited.

Try: "What did I ask about in my Chemistry textbook?"
get_study_statsStudy Stats

Returns total hours studied, completed sessions, current streak, and best streak.

Try: "How many hours have I studied in total?"
search_communityCommunity Q&A

Searches community questions by topic across title, body, and tags.

Try: "Find community questions about calculus integration."
Example prompts
“Plan my revision for today”

Claude reads your weekly tasks and suggests a session order based on what's due.

“I have 3 hours tonight — what should I study?”

Combines your study plan, stats, and resource folders to prioritise.

“What resources do I have for my weakest subject?”

Lists all saved links for the subject with the least study time.

“Summarise what I covered in my last PDF session”

Reads the most recent PDF chat history and condenses the key points.

“How is my study streak going?”

Pulls your current and best streaks and total hours from your stats.

“Find community explanations for photosynthesis”

Searches Q&A for matching questions and surfaces the top answers.

Tips
You do not need to name a specific tool — Claude picks the right one automatically based on your question.
Ask follow-up questions freely. Claude keeps context so "now focus on week 2" works after reading your plan.
Combine tools in one question: "Show me my Biology resources and any community questions on cell division."
If a tool returns no data, check that your Study Cue backend is running and MongoDB is reachable.
The config file path varies by OS — see the config paths table in the sidebar if Claude Desktop does not pick up the server.
What is MCP?

The Model Context Protocol (MCP) is an open standard by Anthropic that lets AI assistants securely read from external data sources. Once connected, Claude can query your Study Cue data the same way it reads a file — no copy-pasting needed.

Requirements
Claude DesktopClaude.ai desktop app
Python 3.12+For running the server
MongoDB runningLocal or Atlas URI
Study Cue backendbackend2 cloned locally
Config file location
macOS
~/Library/Application Support/Claude/claude_desktop_config.json
Windows
%APPDATA%\Claude\claude_desktop_config.json