Doctah-MCP

You set up Doctah-MCP to bridge AI assistants with Arknights data, letting models search operators and enemies and retrieve content in a clean, structured format. This MCP server exposes concise, machine-friendly outputs that you can query from your AI workflows.

How to use

You connect your MCP client to the Doctah-MCP server to perform searches and list operations. Use the provided tools to query operator details, fetch enemy data, and discover operators or enemies via fuzzy matching. The output is formatted in clean Markdown for easy rendering in AI chat and other interfaces.

Typical usage patterns include querying a specific operator by name to retrieve skills and talents, requesting level progression data for enemies, and listing items that match a search pattern. Each response includes a structured payload that your client can parse and present to users or further process in your AI workflow.

How to install

Prerequisites: Python is required to run the MCP server locally. You also need a working Python environment with network access to install dependencies.

Step 1: Clone the project repository and install in editable mode.

git clone https://github.com/TonybotNi/Doctah-MCP.git
cd doctah-mcp
pip install -e .

Step 2: Verify the installation and available commands.

doctah-mcp --help

python -m doctah_mcp.server --help

Step 3: For development, install with development dependencies.

# Clone and set up development environment
git clone https://github.com/TonybotNi/Doctah-MCP.git
cd doctah-mcp

# Install with test dependencies
pip install -e ".[dev]"

Other important topics

MCP integration requires configuring a client to load the Doctah-MCP server. You can connect via a global command, a Python module invocation, or a full path to the Python executable, depending on your environment.

Environment variables can control server behavior. A common variable is LOG_LEVEL to set the verbosity of logs.

If you need help with running locally, ensure your working directory points to the downloaded project folder when using the full path method.

Configuration and troubleshooting

configuring clients: You can set up a client to start the MCP server and point it to the correct entry point. Multiple configurations are supported to accommodate different environments (global command, Python module, or full path).

Common troubleshooting steps include ensuring the command is resolvable in your shell, confirming the working directory for the full-path method, and verifying that the required dependencies are installed in your environment.

Available tools

Operator Search

Query detailed operator information including skills, talents, and attributes.

Enemy Search

Retrieve comprehensive enemy data with level progression and attributes.

List Operators

Find operators that match a given pattern using fuzzy search.

List Enemies

Find enemies that match a given pattern using fuzzy search.