Process Mining MCP Server
A tool that integrates process mining capabilities with Claude AI, allowing users to analyze event log data from PostgreSQL databases through natural language interfaces in Claude Desktop or Cursor.
README Documentation
Project Launch Instructions
IMPORTANT: The project was run on a Windows system.
Steps to follow
-
Install PostgreSQL, create a database and a table with any name.
Access credentials (database name, user, password, table name, etc.) should be entered in the.envfile.
There is a sample.env.examplefile in the repository — use it as a template to fill in your actual.envfile. -
Create an account on Anthropic and add funds to it.
Otherwise, I can share myAPI_KEYvia email. -
The repository contains example
.batlaunch files with the_examplesuffix.
Create your own.batfiles (without_examplein the name), filling in the appropriate project paths. -
Create a virtual environment and install the required libraries according to
requirements.txt:
pip install -r requirements.txt
- Download Claude Desktop.
Go toFile -> Settings -> Developer -> Edit Config.
Paste the contents of theclaude_desktop_config_example.json, file there, filling in the paths to your.batfiles.
Make sure to use the correct path format.
-
Extract the file
BPI Challenge 2017.xes.gzinto thedatasetsfolder. -
Run the data import script:
python import_data.py
The script should import the event log data into the database.
- Restart Claude Desktop.
After launching the tool, the tools should be visible below the input field (two sliders labeled "search and tools"). There you can enable or disable the available tools.
Optional: Configuring MCP in Cursor
If you are using Cursor with MCP tools:
- Go to
File -> Preferences -> Cursor Settings. - Go to
MCP Toolstab. - Press
New MCP Server buttonand paste previously filledclaude_desktop_config.jsonfile contents, adjusting the paths to your.batfiles as needed. - Save the configuration. Your MCP tools should now be available inside Cursor.