Mousetail

Mousetail

The simplest and most stable MCP server for Anki, enabling LLMs to perform core Anki operations (CRUD) directly without addons.

Category
访问服务器

README

<div align="center"> <img src="https://github.com/listfold/mousetail/blob/main/mousetail_transparent.png?raw=true" alt="Mousetail Logo" width="200"/> </div>

The simplest and most stable MCP Server for Anki

UV Python 3.10+ MCP License: MIT

Mousetail's goal is to be the simplest and most reliable way of connecting Anki to an LLM. It does not require any addons, just have anki installed and an LLM you'd like to connect to your decks.

Running the server is as simple as:

uvx mousetail

For detailed instructions on integrating with Claude Code, Claude Desktop, and other LLM tools:

Usage Guide

API Reference

Features

  • Minimal - supports core anki operations: create, read, update and delete
  • Stable - works directly with anki's stable pylib api, no addons or deps
  • Simple - no configuration required, automatically discovers your Anki collections

Use Cases

Selectively commit what you learn in conversation with an LLM to memory

"Create a deck based on our conversation so I don't forget critical details"

Use an LLM to interact with your deck

"Create a card - what is a coeffcient?, create a card - id the first coefficient in this polynomial"

Use an LLM to cleanup a deck

"Check my french vocab deck for correctness and correct any mistakes, typos or errors"

Syncing

Mousetail supports synchronizing your Anki collection with AnkiWeb or a self-hosted sync server. This allows you to keep your collection in sync across devices.

Quick Start

Sync with AnkiWeb (saves credentials for future syncs):

> "Save my AnkiWeb credentials: username is myuser@example.com and password is mypassword"
> "Sync my collection with AnkiWeb"

Credential Management

Mousetail stores credentials securely in your system's credential manager:

  • macOS: Keychain
  • Windows: Credential Manager
  • Linux: Secret Service (GNOME Keyring, KWallet, etc.)

Available credential tools:

  • save_sync_credentials - Save username/password securely
  • load_sync_credentials - Load saved credentials
  • delete_sync_credentials - Remove saved credentials

Sync Options

The sync_collection tool supports:

  • AnkiWeb sync (default) - Leave endpoint empty
  • Self-hosted servers - Provide custom endpoint URL (e.g., https://sync.example.com)
  • Media sync - Enabled by default, includes images and audio files
  • Collection-only sync - Set sync_media: false to skip media

Examples

First-time setup with AnkiWeb:

> "Save my sync credentials for AnkiWeb - username: user@example.com, password: mypass123"
> "Sync my collection"

Using a self-hosted server:

> "Save my sync credentials - username: john, password: secret, endpoint: https://sync.myserver.com"
> "Sync my collection"

One-time sync without saving credentials:

> "Sync my collection with AnkiWeb using username user@example.com and password mypass123"

Collection-only sync (skip media):

> "Sync my collection but don't sync media files"

Configuration

You can set a default sync endpoint in config.json:

{
  "sync": {
    "endpoint": "https://sync.example.com"
  }
}

Leave endpoint as null to use AnkiWeb by default.

Important Notes

  • Close Anki first: Sync will fail if the Anki application is running
  • Media sync: Media sync is slower and uses more bandwidth but ensures images/audio are synced
  • Conflicts: If conflicts occur, try syncing from Anki desktop first to resolve them
  • Security: Credentials are never stored in plain text - they're kept in your system's secure credential storage

Self-Hosted Sync Server Setup

To sync with your own server:

  1. Set up an Anki sync server
  2. Configure credentials on the server (using environment variables like SYNC_USER1=user:password)
  3. Save your credentials in Mousetail with the server endpoint
  4. Sync normally

For detailed sync server setup, see the official Anki documentation.

Important Notes

How Collections Are Accessed

The MCP server finds Anki collections at their standard locations:

  • macOS: ~/Library/Application Support/Anki2/[Profile]/collection.anki2
  • Linux: ~/.local/share/Anki2/[Profile]/collection.anki2
  • Windows: %APPDATA%\Anki2\[Profile]\collection.anki2

You don't need to configure paths - the server automatically discovers available collections, this can be customized using configuration.

Configuration

The server can be customized through a config.json file. See the Usage Guide for configuration options.

Development

Local Development Setup

To develop and test Mousetail locally with Claude Code:

  1. Clone the repository:

    git clone https://github.com/listfold/mousetail.git
    cd mousetail
    
  2. Install dependencies:

    uv sync
    
  3. Configure Claude Code: The project includes a .mcp.json file that configures the MCP server for local development:

    {
      "mcpServers": {
        "mousetail": {
          "type": "stdio",
          "command": "uv",
          "args": ["run", "python", "-m", "mousetail.mcp.stdio_server"],
          "env": {
            "PYTHONUNBUFFERED": "1"
          }
        }
      }
    }
    
  4. Restart Claude Code: After the configuration is in place, restart Claude Code to load the MCP server.

  5. Verify the server:

    • Use /context in Claude Code to see available MCP tools
    • The mousetail server should appear with all available tools (list_collections, create_note, sync_collection, etc.)
  6. Testing sync functionality:

    • Close the Anki desktop application before testing
    • Test credential management: save_sync_credentials, load_sync_credentials, delete_sync_credentials
    • Test sync: sync_collection with your AnkiWeb credentials or self-hosted server

Core goals

Mousetail was written because all the existing MCP Anki tools depend on the AnkiConnect addon.

AnkiConnect is a HTTP server for Anki, it was originally created to support connecting browser extensions like yomichan to Anki. For MCP development, it is not necessary and introduces issues:

  • introduces complexity e.g. a dedicated HTTP server for Anki occupies a port
  • introduces risk e.g. if the AnkiConnect API changes or has a bug the MCP tool will break
  • introduces an extra step e.g. all current MCP tools require installing the AnkiConnect addon

Mousetail has a much simpler approach. It integrates directly with Anki's pylib. This is a stable API that's part of Anki's core, it therefore is not subject to arbitrary or frequent change, and does not require any 3rd-party addons.

Because it prioritizes simplicity, mousetail will remain more stable than the alternatives. The tradeoff is that Mousetail will never integrate with the Anki UI. It is also reasonable to assume that Mousetail will only ever work with colocated (same system) LLM tools and Anki decks.

Building Documentation

The project uses Sphinx with the Furo theme to generate documentation from Python docstrings.

  1. Install documentation dependencies:

    uv pip install ".[docs]"
    
  2. Build the documentation:

    uv run python -m sphinx -b html docs docs/_build/html
    
  3. View the documentation: Open docs/_build/html/index.html in your browser.

The documentation is automatically built and deployed to GitHub Pages on every push to the main branch.

License

MIT License - see LICENSE file for details.

推荐服务器

Baidu Map

Baidu Map

百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。

官方
精选
JavaScript
Playwright MCP Server

Playwright MCP Server

一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。

官方
精选
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。

官方
精选
本地
TypeScript
VeyraX

VeyraX

一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。

官方
精选
本地
graphlit-mcp-server

graphlit-mcp-server

模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。

官方
精选
TypeScript
Kagi MCP Server

Kagi MCP Server

一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

Exa MCP Server

模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。

官方
精选