v0.3.0 MIT License Python 3.9+

Stata CLI Is All Reg Monkeys Need

A command-line interface for Stata via PyStata — built for humans and AI Agents. Run code, view data, export graphs, all with sub-second execution.

pip install stata-cli GitHub

$ stata-cli daemon start

Daemon started (PID 42891)

$ stata-cli run "sysuse auto, clear

regress price mpg weight"

Source | SS df MS

-----------+------------------------------

Model | 518727240 2 259363620

Residual | 117320986 71 1652408.3

price | Coefficient Std. err.

-----------+-------------------------

mpg | -49.51222 86.15604

weight | 1.746559 .6413538

Execution time: 0.085s

Built for Modern Workflows

Everything you need to run Stata from the terminal or integrate with AI agents.

🤖

Agent-Native Design

Structured JSON output, exit codes, and SKILL.md definition. AI Agents operate Stata with zero extra setup.

⚡

Sub-Second Execution

Daemon mode keeps PyStata alive in background. Startup drops from ~2-3s to ~85ms — a 35x speedup.

🎯

Full Coverage

Run code, .do files, view data, browse help, export graphs, inspect matrices — one tool for everything.

📊

AI-Friendly Output

Compact mode, token limit management, JSON responses, and graph auto-naming — optimized for tool-use.

🔓

Open Source

MIT license. No barriers. pip install and you're running Stata from the command line in seconds.

🚀

2-Step Setup

Auto-detects your Stata installation. From install to first command in under 30 seconds.

Quick Start

Up and running in seconds — for humans and AI agents alike.

🧑‍💻 Human Users

# Install
pip install stata-cli

# Verify Stata is detected
stata-cli detect

# Run your first command
stata-cli run "display 1+1"

# Start daemon for 35x faster execution
stata-cli daemon start
stata-cli run "sysuse auto, clear"

🤖 AI Agents

                    
                    # Install & start daemon
pip install stata-cli
stata-cli daemon start

# Run analysis
stata-cli run "sysuse auto, clear
regress price mpg weight"

# Get structured results
stata-cli --json run "summarize price"
stata-cli return e
stata-cli matrix e(b)
                

Commands

14 commands covering the full Stata workflow.

Command	Description	Example
`run`	Execute inline Stata code or pipe from stdin	`stata-cli run "regress y x"`
`do`	Execute .do files with /// continuation	`stata-cli do analysis.do`
`data`	View current dataset as JSON	`stata-cli data --if "price>5000"`
`return`	Retrieve r(), e(), s() stored results	`stata-cli return e`
`vars`	Inspect variable metadata	`stata-cli vars price mpg`
`matrix`	Read Stata matrices as JSON	`stata-cli matrix e(b)`
`labels`	List and inspect value labels	`stata-cli labels origin`
`macro`	Get/set Stata macros	`stata-cli macro get "c(N)"`
`frame`	List Stata frames	`stata-cli frame`
`skill`	Built-in Stata reference library (57 topics)	`stata-cli skill regression`
`help`	Browse Stata help topics	`stata-cli help regress`
`stop`	Interrupt running command	`stata-cli stop`
`detect`	Find Stata installation path	`stata-cli detect`
`daemon`	Manage background daemon	`stata-cli daemon start`

Daemon Mode Performance

Keep PyStata alive in the background. No more waiting for initialization.

~2.5s

Without daemon

→

~85ms

With daemon

35x faster — daemon keeps PyStata initialized via Unix socket