ankimcp/anki-mcp-server

What it solves

This project provides a bridge between AI assistants and Anki, a popular spaced repetition flashcard application. It allows users to interact with their study materials using natural language, enabling AI to act as a private tutor that can explain concepts, create and edit notes, and manage decks dynamically during study sessions.

How it works

The project is implemented as a Model Context Protocol (MCP) server. It connects to Anki via the AnkiConnect plugin. It can be run in three primary modes:

• STDIO: Used by local desktop AI clients (like Claude Desktop or Cursor) to communicate with the server as a subprocess.
• HTTP: Runs as a local web server for web-based AI tools.
• Tunnel: Creates a secure, authenticated WebSocket connection to a managed tunnel service, allowing cloud-hosted AI assistants (like ChatGPT or Claude.ai) to access a local Anki installation.

Who it’s for

Students and lifelong learners who use Anki for spaced repetition and want to integrate AI-driven tutoring, automated card creation, and natural language management of their flashcard collection.

Highlights

• Comprehensive Toolset: Exposes 42 MCP tools, including 31 essential tools for review, deck, note, tag, and media management, and 11 GUI tools to drive the Anki desktop interface.
• Flexible Integration: Supports multiple transport methods (including a secure tunnel for remote AI) to connect with various AI clients.
• Dynamic Card Creation: Ability to create custom note types (models) with CSS styling for specific needs, such as right-to-left (RTL) languages.
• Media Handling: Efficiently manages media uploads via file paths or URLs to avoid token inefficiency.
• Study Automation: Enables AI to sync with AnkiWeb, fetch due cards, present them and record ratings based on user performance.