Give your AI agents
real shell power.

A simple, secure HTTP server that lets AI agents execute commands on your machine and get structured results back.

pip install AI-SystemAssist

View on PyPI

Trusted by AI agents worldwide

CAPABILITIES

Built for agents.
Built for humans.

Read full documentation

Execute anything

Run any shell command. Pipes, redirects, background processes — everything works.

curl -X POST -d 'ls -la | grep py' ...

Structured output

Get clean JSON back with stdout, stderr, return code, and execution duration.

{ "stdout": "...", "returncode": 0, "success": true }

Secure by default

Token-based auth. Binds to localhost by default. Supports full root access when needed.

X-Agent-Token

Bearer

System visibility

Perfect for agents that need to understand disk space, processes, network, and environment.

Privilege control

Run as regular user or with full sudo/root privileges. Your choice.

Works everywhere

curl, wget, Python, Node.js, or directly from your LLM agent. No special client needed.

curl

wget

Python

+ more

HOW IT WORKS

Simple. Powerful. Agent-ready.

01

Install & start

Install the package and start the server with a single command. Optionally run with sudo for full access.

02

Send commands

Your agent (or curl, Python, etc.) POSTs commands to the server with the auth token.

03

Get results

Receive clean JSON containing stdout, stderr, return code, and execution time. Ready for your agent to reason over.

Ready to give your agents real power?