What can you do with this server?

This server enables interaction with Stata for data analysis and management through five key capabilities: * **Read Log Files:** Retrieve and return the content of specified log files * **Get Data Information:** Obtain descriptive statistics for data files, with optional encoding and variable list specifications * **Write Dofiles:** Create new Stata dofiles with provided content * **Append Dofiles:** Add Stata code to existing dofiles or create new ones if they don't exist * **Execute Stata Code:** Run Stata commands by executing specified dofiles through Stata

Which integrations are available for this server?

Supports running Stata statistical analysis on [macOS](/mcp/servers/integrations/macos) systems, with Windows support planned for future releases.

Stata-MCP by SepineTam

PyPI version PyPI Downloads License: Apache 2.0 Issue Ask DeepWiki

News:

Use Stata-MCP in Claude Code, look here
Try to use agent mode as tool? Now it is supported more easily here.
Want to evaluate your LLM? Look here.
Update StataFinder, but it is not stable, please config your STATA_CLI into your environment.

Finding our newest research? Click here or visit reports website.

Looking for others?
STOP: StataMCP-Team Opendata Project 📊, we have open-sourced a comprehensive dataset collection for social science research, aiming to enable the future of AI-driven and data-powered research paradigms.
Trace DID: If you want to fetch the newest information about DID (Difference-in-Difference), click here. Now there is a Chinese translation by Sepine Tam and StataMCP-Team 🎉
Jupyter Lab Usage (Important: Stata 17+) here
NBER-MCP & AER-MCP 🔧 under construction
Econometrics-Agent
TexIV: A machine learning-driven framework that transforms text data into usable variables for empirical research using advanced NLP and ML techniques
A VScode or Cursor integrated here. Confused it? 💡 Difference

💡 Quick Start

Use Stata-MCP in Claude Code

We can use Stata-MCP in Claude Code as its prefect agentic ability.

Before using it, please make sure you have ever install Claude Code, if you don't know how to install it, visit on GitHub

You can open your terminal and cd to your working directory, and run:

claude mcp add stata-mcp --env STATA_MCP_CWD=$(pwd) -- uvx stata-mcp

I am not sure whether it works on Windows, as I do not have a Windows device for test it.

Then, you can use Stata-MCP in Claude Code. Here are some scenarios for using it:

Paper Replication: Replicate empirical studies from economics papers
Quick Hypothesis Testing: Validate economic hypotheses through regression analysis
Stata Learning Assistant: Learn econometrics with step-by-step Stata explanations
Code Organization: Review and optimize existing Stata do-files
Result Interpretation: Understand complex statistical outputs and regression results

Agent Mode

The details of agent mode find here.

git clone https://github.com/sepinetam/stata-mcp.git cd stata-mcp uv sync uv pip install -e . stata-mcp --version # for test whether stata-mcp is installed successfully. stata-mcp --agent # now you have enjoy your stata-mcp agent mode.

or you can directly use it with uvx:

uvx stata-mcp --version # for test whether it could be used on your computer. uvx stata-mcp --agent

You can edit the task in agent_examples/openai/main.py for variable model_instructions and task_message, click me #L37 and #L68

Agent as Tool

If you want to use a Stata-Agent in another agent, here is a simple example:

import asyncio from agents import Agent, Runner from stata_mcp.agent_as.agent_as_tool import StataAgent # init stata agent and set as tool stata_agent = StataAgent() sa_tool = stata_agent.as_tool() # Create main Agent agent = Agent( name="Assistant", instructions="You are a helpful assistant", tools=[sa_tool], ) # Then run the agent as usual. async def main(task: str, max_turns: int = 30): result = await Runner.run(agent, input=task, max_turns=max_turns) return result if __name__ == "__main__": econ_task = "Use Stata default data to find out the relationship between mpg and price." asyncio.run(main(econ_task))

AI Chat-Bot Client Mode

Standard config requires: please make sure the stata is installed at the default path, and the stata cli (for macOS and Linux) exists.

The standard config json as follows, you can DIY your config via add envs.

{ "mcpServers": { "stata-mcp": { "command": "uvx", "args": [ "stata-mcp" ] } } }

For more detailed usage information, visit the Usage guide.

And some advanced usage, visit the Advanced guide

Prerequisites

uv - Package installer and virtual environment manager
Claude, Cline, ChatWise, or other LLM service
Stata License
Your API-KEY from LLM

Notes:
If you are located in China, a short uv usage document you can find here.
Claude is the best choice for Stata-MCP, for Chinese, I recommend to use DeepSeek as your model provider as it is cheap and powerful, also the score is highest in China provider, if you are increased in it, visit the report How to use StataMCP improve your social science research.

Installation

For the new version, you don't need to install the stata-mcp package again, you can just use the following command to check whether your computer can use stata-mcp.

uvx stata-mcp --usable uvx stata-mcp --version

If you want to use it locally, you can install it via pip or download the source code.

Download via pip

pip install stata-mcp

Download source code and compile

git clone https://github.com/sepinetam/stata-mcp.git cd stata-mcp uv build

Then you can find the compiled stata-mcp binary in the dist directory. You can use it directly or add it to your PATH.

For example:

uvx /path/to/your/whl/stata_mcp-1.13.0-py3-non-any.whl # here is the wheel file name, you can change it to your version

Related MCP server: MCP-MultilspyLSP

📝 Documentation

For more detailed usage information, visit the Usage guide.
Advanced Usage, visit the Advanced
Some questions, visit the Questions
Difference with Stata-MCP@hanlulong, visit the Difference

💡 Questions

🚀 Roadmap

macOS support
Windows support
Additional LLM integrations (With a new webUI)
Performance optimizations (Via prompt and context engineering)

For more information, refer to the Statement.

🐛 Report Issues

If you encounter any bugs or have feature requests, please open an issue.

📄 License

Apache License 2.0

📚 Citation

If you use Stata-MCP in your research, please cite this repository using one of the following formats:

BibTeX

@software{sepinetam2025stata, author = {Song Tan}, title = {Stata-MCP: Let LLM help you achieve your regression analysis with Stata}, year = {2025}, url = {https://github.com/sepinetam/stata-mcp}, version = {1.13.0} }

APA

Song Tan. (2025). Stata-MCP: Let LLM help you achieve your regression analysis with Stata (Version 1.13.0) [Computer software]. https://github.com/sepinetam/stata-mcp

Chicago

Song Tan. 2025. "Stata-MCP: Let LLM help you achieve your regression analysis with Stata." Version 1.13.0. https://github.com/sepinetam/stata-mcp.

📬 Contact

Email: sepinetam@gmail.com

Or contribute directly by submitting a Pull Request! We welcome contributions of all kinds, from bug fixes to new features.

❤️ Acknowledgements

The author sincerely thanks the Stata official team for their support and the Stata License for authorizing the test development.

📃 Statement

The Stata referred to in this project is the commercial software Stata developed by StataCorp LLC. This project is not affiliated with, endorsed by, or sponsored by StataCorp LLC. This project does not include the Stata software or any installation packages; users must obtain and install a validly licensed copy of Stata from StataCorp. This project is licensed under Apache-2.0. The project maintainers accept no liability for any loss or damage arising from the use of this project or from actions related to Stata.

More information: refer to the Chinese version at [source/docs/README/cn/README.md]; in case of any conflict, the Chinese version shall prevail.

Stata-MCP