Anki Mcp

io.github.arielbkPublished by GitHub account arielbkv0.3.2Updated Oct 24, 2026

MCP server for integrating with Anki flashcards through conversational AI

Overview

AI-generated overview

Local MCP server that connects an AI assistant to your Anki flashcard collection through AnkiConnect for quizzing, card creation, organization and study…

What you can do
  • Create, update, delete and search flashcards, including bulk edits and tagging
  • Run quiz sessions, manage card scheduling and find cards due for review
  • Create and organize decks, move cards between them and adjust deck settings
  • Read study statistics, review history and retention patterns
  • Customize note types, fields, templates and card styling
  • Handle sync, media files, imports and exports, and retrieve images or audio from the collection
When to use it
Useful when you study with Anki and want an assistant to quiz you on specific decks or cards, generate new cards from notes or documents, bulk-tag or reorganize cards, or report on your study patterns and weak areas.
Requirements
Runs locally as an npm package (Node.js 18 or newer) on desktop. Anki must be installed and running with the AnkiConnect add-on (version 25.2.25.0 or newer recommended); the server connects to AnkiConnect on localhost:8765 by default, adjustable via ANKI_CONNECT_HOST and ANKI_CONNECT_PORT. Anki must be started manually.
Before you install
The assistant can change your collection: it can create, update and delete cards, move cards between decks, retag thousands of cards in bulk, and run sync, import and export operations. Point it at a collection you can restore from a backup, and review bulk or delete requests before approving them.

Installation

In SourceWeft

  1. Open Anki Mcp in the dashboard and add it to a workspace.
  2. Enable the server for the chats that should use its tools.

Desktop only via STDIO. STDIO servers start a local process, so they need the SourceWeft desktop host.

Other MCP clients

Follow the launch instructions in the repository.

README

🧠 Anki MCP

[NPM Package anki-mcp] [License: MIT]
[Anki MCP server]

Transform your Anki flashcard experience with AI! This MCP (Model Context Protocol) server bridges your Anki flashcard collection with AI assistants like Claude, enabling natural conversations about your study materials.

🚀 What Can You Do?

Imagine having an AI tutor that knows your entire flashcard collection and can:

📚 Interactive Learning

  • "Quiz me on Japanese vocabulary I haven't seen in 3 days"
  • "Test me on challenging cards from my medical deck"
  • "Show me cards I've been struggling with recently"

✨ Smart Content Creation

  • "Create flashcards about photosynthesis with diagrams"
  • "Turn this PDF chapter into spaced repetition cards"
  • "Generate cloze deletion cards from my lecture notes"

🔍 Powerful Analytics

  • "Which topics am I struggling with most?"
  • "Show me my study patterns for the last month"
  • "What's my retention rate for different card types?"

🎯 Bulk Operations

  • "Tag all my chemistry cards with 'exam-prep'"
  • "Move cards with low retention to an intensive review deck"
  • "Find and fix duplicate cards across my decks"

🎯 Features

  • 🎯 Smart Tool Design: Intuitive, high-level tools designed specifically for AI agents
  • 💬 Natural Conversations: Talk to Claude about your flashcards like you would a study buddy
  • ⚡ Lightning Fast: Efficient operations that don't bog down your AI assistant
  • 📦 Bulk Operations: Update thousands of cards at once with simple commands
  • 📊 Study Insights: Deep analytics on your learning progress and patterns
  • 🎨 Rich Media: Full support for images, audio, and other media in your cards
  • 🔍 Powerful Search: Leverage Anki's advanced search syntax through natural language
  • 🔒 Rock Solid: Fully typed TypeScript with comprehensive error handling

🛠 Technical Stack

  • Node.js: Runtime environment (18+ required)
  • TypeScript: Type safety and developer experience
  • MCP SDK: Model Context Protocol implementation
  • yanki-connect: 🙏 Fully-typed, isomorphic AnkiConnect API client
  • tsup: Fast TypeScript bundler
  • PNPM: Efficient package manager

Special thanks to the yanki-connect library for providing the robust AnkiConnect interface that powers this integration!

📋 Prerequisites

1. Install Anki

Download and install Anki if you haven't already.

2. Install AnkiConnect Plugin

  1. Open Anki
  2. Go to Tools → Add-ons
  3. Click Get Add-ons...
  4. Enter code: 2055492159
  5. Restart Anki

📋 Version Requirements: This server requires AnkiConnect version 25.2.25.0 or newer (released 2025-02-25) for full compatibility. Most features work with older versions, but we recommend updating for the best experience.

3. Configure AnkiConnect (Optional)

AnkiConnect works out of the box, but you can customize settings:

  1. Go to Tools → Add-ons
  2. Select AnkiConnect and click Config
  3. Default settings should work fine for most users

🚀 Quick Start

Option 1: Use with Claude Desktop (Recommended)

  1. Install the package

    bash
    npm install -g @arielbk/anki-mcp
  2. Configure Claude Desktop

    Edit your Claude Desktop configuration file:

    macOS: ~/Library/Application Support/Claude/claude_desktop_config.json Windows: %APPDATA%\Claude\claude_desktop_config.json

    Add this configuration:

    json
    {  "mcpServers": {    "anki": {      "command": "anki-mcp",      "args": []    }  }}
  3. Restart Claude Desktop

  4. Start using it!

    • Make sure Anki is running with AnkiConnect enabled
    • Ask Claude: "Show me my Anki decks" or "Quiz me with 5 cards"

💡 Pro Tip: You can test your setup by asking Claude simple questions like "How many Anki decks do I have?" or "What's in my largest deck?"

Option 2: Development Setup

  1. Clone the repository

    bash
    git clone https://github.com/arielbk/anki-mcp.gitcd anki-mcp
  2. Install dependencies

    bash
    pnpm install
  3. Build the project

    bash
    pnpm build
  4. Test the server

    bash
    # Use the MCP inspector to test functionalitypnpm inspect

Option 3: Use with Other MCP Clients

This server works with any MCP-compatible client. You can also use it with:

  • MCP Inspector: Interactive debugging tool
  • Cline: VS Code extension for AI coding
  • Custom MCP clients: Build your own using the MCP SDK

For stdio transport, run:

bash
node dist/index.js

For localhost HTTP (Streamable HTTP) transport, run:

bash
node dist/index.js --http --host=127.0.0.1 --port=3000# The MCP endpoint will be at http://127.0.0.1:3000/mcp

Clients that support Streamable HTTP can connect directly. Legacy SSE clients may still work via client-side fallback.

🎮 Usage Examples

Once configured with Claude Desktop, you can have natural conversations like:

📖 Study Sessions

You: "I have a Japanese exam tomorrow. Can you quiz me on Hiragana cards I haven't reviewed in a week?"
Claude: I'll find your Hiragana cards that need review and start a quiz session...

📝 Content Creation

You: "Create flashcards for the major battles of World War II with dates and significance"
Claude: I'll create comprehensive flashcards covering the key WWII battles...

📊 Study Analytics

You: "How am I doing with my medical terminology deck? Show me my weak areas."
Claude: Let me analyze your performance on the medical terminology deck...

🏷️ Organization

You: "Tag all cards about cardiovascular system with 'cardiology' and move them to my MCAT prep deck"
Claude: I'll help you organize those cards...

🔧 Available Tools

The server provides 7 intuitive tools that cover everything you need:

📝 Manage Flashcards

Create, update, delete, and search your flashcards. Perfect for content creation, bulk updates, and organization.

🎯 Study Sessions

Run interactive quiz sessions, manage card scheduling, and track which cards need review.

📚 Manage Decks

Create and organize decks, move cards around, and configure deck settings.

📊 Get Analytics

Dive into your study statistics, review history, and learning patterns.

🎨 Manage Models

Customize note types, add fields, modify templates, and style your cards.

🔧 Anki Operations

Handle sync, media files, imports/exports, and other utility operations.

🖼️ Get Media File

Retrieve media files (images, audio, etc.) from your Anki collection. AI assistants can analyze images directly (though they won't display in the UI).

💡 Pro Tip: Many operations support pagination to prevent overwhelming your context window. Use limit and offset parameters when working with large result sets.

🛠 Development

bash
# Run in development mode with auto-reloadingpnpm dev
# Lint the codepnpm lint
# Format the codepnpm format
# Test with MCP inspectorpnpm inspect

🔧 Configuration

AnkiConnect Settings

The server connects to AnkiConnect on localhost:8765 by default. If you've customized your AnkiConnect configuration, you may need to adjust the connection settings.

Important: Make sure Anki is running and AnkiConnect is installed before using the MCP server. The server will fail to connect if AnkiConnect is not accessible.

💡 Note: The underlying yanki-connect library supports auto-launching Anki on macOS, but this feature is not currently exposed in the MCP server configuration. You'll need to start Anki manually.

Claude Desktop Advanced Configuration

For advanced users, you can pass additional parameters:

json
{  "mcpServers": {    "anki": {      "command": "anki-mcp",      "args": ["--verbose"],      "env": {        "ANKI_CONNECT_HOST": "localhost",        "ANKI_CONNECT_PORT": "8765"      }    }  }}

Troubleshooting

Common Issues:

  • "Failed to connect to AnkiConnect": Ensure Anki is running and AnkiConnect plugin is installed
  • "No decks found": Make sure you have at least one deck in Anki
  • Claude doesn't respond: Restart Claude Desktop after configuration changes
  • Permission errors: Check that the global npm install worked correctly with npm list -g @arielbk/anki-mcp

🤝 Contributing

We welcome contributions! Please:

  1. Fork the repository
  2. Create a feature branch
  3. Make your changes
  4. Add tests if applicable
  5. Submit a pull request

📜 License

MIT - see LICENSE file for details.

🙏 Acknowledgments


Transform your flashcard experience with AI-powered conversations! 🚀

Source: README.md at commit de77ac5

Tools

0
SourceWeft has not started this package in its sandbox yet, so its tools are not known.

Version history

1
  1. v0.3.2LatestSep 16, 2026

Score

50% of the applicable checks met

How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.

  • Reached in the last 30 days25 pts

    Keep the endpoint answering the MCP handshake; SourceWeft probes it every 30 days.

  • Auth declared10 pts

    Declare the credential headers in the registry entry, or keep the endpoint answering the probe so its auth can be classified.

  • Tools listed15 pts

    Answer tools/list without a credential, or ship a package that starts in a sandbox without one.

  • DescriptionRequired15 pts
  • Logo10 pts
  • Documentation15 pts
  • License10 pts