mcp -anki

This MCP server integrates Anki with the MCP framework using uvx to run the mcp-server-anki helper. It relies on Anki Desktop, Anki Connect, and uv to bridge your workflow into MCP-compatible commands. The server exposes two primary ways to configure and interact with Anki:

• Cursor mode: add the anki server to your .mcp.json under mcpServers so MCP can forward commands to the uvx mcp-server-anki instance. This mode enables standard MCP cursor-based interactions, allowing you to issue updates that reflect in Anki decks and notes via the Anki Connect bridge.
• Chatwise mode: configure a UI-based integration where the MCP system runs uvx mcp-server-anki from the chat/command surface. This lets you trigger Anki-related actions through a simple chat-like interface, using the ID "Anki" and the command "uvx mcp-server-anki".

Once set up, you can perform typical tasks such as creating or updating notes, adding cards, or syncing with Anki via Anki Connect-friendly commands exposed by the mcp-server-anki implementation. The server is intended to work with an Anki Desktop installation and the Anki Connect plugin, and it should be accessible through the uvx runtime as configured above.

Prerequisites

• Anki Desktop: Install from https://apps.ankiweb.net/
• Anki Connect: Install extension from https://ankiweb.net/shared/info/2055492159
• uv (uvx): Install UV runtime (see https://docs.astral.sh/uv/getting-started/installation/)

Installation steps

1. Install prerequisites

• Install Anki Desktop on your operating system.
• Install Anki Connect in Anki (via the Add-ons manager) to enable programmatic access.
• Install UV (the uv runtime) for your platform.

2. Prepare the MCP server files

• Clone or download the mcp-server-anki project files into a local development directory.
• Ensure you have a Python environment available if you plan to run the Python-based server (as shown in the local development example).

3. Run the MCP server via UVX

• In your environment, start the mcp-server-anki service with uvx, as described in the README usage:

uvx mcp-server-anki

4. Optional: configure MCP client

• Update your .mcp.json to include the Anki MCP server configuration (see the example in the README).
• Ensure your MCP client is aware of the Anki integration and can send/receive commands through the uvx bridge.

5. Validate

• Confirm Anki Desktop is running with Anki Connect available.
• Verify that commands issued through MCP reach Anki via Anki Connect and that notes/cards update as expected.

Tips and considerations:

• Ensure Anki Desktop and Anki Connect are running and accessible before issuing MCP commands.
• The cursor configuration uses uvx with the mcp-server-anki package name, so the working directory and PATH should allow uvx to locate the command.
• In chatwise mode, the configured command is uvx mcp-server-anki with ID "Anki". Use this through your MCP-enabled chat or UI to trigger actions.
• If you run into connection issues, verify that Anki Connect is enabled in Anki and that the port/endpoint it uses is reachable by the uvx runtime.
• This server is built to align with the Anki ecosystem; if you upgrade Anki or Anki Connect, re-check compatibility with the mcp-server-anki implementation.
• For local development, the provided example shows running server.py under a specific directory; adapt paths to your environment as needed.