Auto Causal Inference MCP server

This MCP server provides an automated causal inference pipeline for banking applications, enabling users to discover causal relationships by simply specifying a treatment and outcome variable.

Installation

To get started with the Auto Causal Inference MCP server, follow these steps:

Prerequisites

• Python 3.10
• Claude Desktop (to run MCP)

Installation Steps

Install the required dependencies:

pip install requirements.txt

Usage

The Auto Causal Inference server can be run in two different modes:

Running with LangGraph

To start the server using LangGraph:

cd agent
python app.py

If you want to test with LangGraph Studio:

langgraph dev

The UI will be available at: https://smith.langchain.com/studio/?baseUrl=http://127.0.0.1:2024

Running with MCP and Claude Desktop

To run the MCP implementation:

cd mcp_agent
python client.py

Example Usage

Input Format

You can ask causal questions about your banking data. For example:

Does offering a promotion increase digital product activation?

Output Components

The system will automatically:

1. Search relevant variables in the database
2. Find causal relationships with CausalNex
3. Identify causal variables
4. Perform Causal Model with DoWhy
5. Find the best estimators with CausalTune
6. Run refutation tests to validate the causal structure

Example Output

The system will provide:

1. Causal Relationships - showing all the causal connections between variables
2. Causal Variables - categorizing variables as confounders, mediators, effect modifiers, etc.
3. Average Treatment Effect (ATE) - calculating the causal impact
4. Model Tuning Results - finding the optimal causal estimator
5. Refutation Test Results - validating the causal structure
6. Summary Analysis - explaining the findings in business terms

Use Cases

The Auto Causal Inference system can answer questions like:

• Does promotion offer increase internet banking activation?
• Do branch visits increase customer engagement?
• Does education level affect income?
• Does channel preference affect internet banking usage?

Available Variables

The system works with the following banking variables:

• age - Customer age
• income - Customer income level
• education - Education level of customer
• branch_visits - Number of times the customer visited a physical branch
• channel_preference - Preferred communication channels
• customer_engagement - Composite metric of interactions
• region_code - Geographic region identifier
• promotion_offer - Whether the customer received a promotion
• activated_ib - Whether the customer activated Internet Banking

How to install this MCP server

For Claude Code

To add this MCP server to Claude Code, run this command in your terminal:

claude mcp add-json "auto-causal-inference" '{"command":"python","args":["mcp_agent/server.py"]}'

See the official Claude Code MCP documentation for more details.

For Cursor

There are two ways to add an MCP server to Cursor. The most common way is to add the server globally in the ~/.cursor/mcp.json file so that it is available in all of your projects.

If you only need the server in a single project, you can add it to the project instead by creating or adding it to the .cursor/mcp.json file.

Adding an MCP server to Cursor globally

To add a global MCP server go to Cursor Settings > Tools & Integrations and click "New MCP Server".

When you click that button the ~/.cursor/mcp.json file will be opened and you can add your server like this:

{
    "mcpServers": {
        "auto-causal-inference": {
            "command": "python",
            "args": [
                "mcp_agent/server.py"
            ]
        }
    }
}

Adding an MCP server to a project

To add an MCP server to a project you can create a new .cursor/mcp.json file or add it to the existing one. This will look exactly the same as the global MCP server example above.

How to use the MCP server

Once the server is installed, you might need to head back to Settings > MCP and click the refresh button.

The Cursor agent will then be able to see the available tools the added MCP server has available and will call them when it needs to.

You can also explicitly ask the agent to use the tool by mentioning the tool name and describing what the function does.

For Claude Desktop

To add this MCP server to Claude Desktop:

1. Find your configuration file:

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

2. Add this to your configuration file:

{
    "mcpServers": {
        "auto-causal-inference": {
            "command": "python",
            "args": [
                "mcp_agent/server.py"
            ]
        }
    }
}

3. Restart Claude Desktop for the changes to take effect