Skip to main content
This example shows the complete developer loop: receive code, create a sandbox, write a file, execute it, return output, and always destroy the sandbox.

​
Prerequisites

  • A running StacyVM server.
  • Docker provider access on the StacyVM host.
  • Python 3.9+.
  • A StacyVM API key if auth is enabled.
The source lives in examples/code-runner-python.

​
Install

cd examples/code-runner-python
python3 -m venv .venv
. .venv/bin/activate
pip install -r requirements.txt

​
Configure

export STACYVM_URL="http://localhost:7423"
export STACYVM_API_KEY="sk_test_YOUR_API_KEY"
export STACYVM_IMAGE="python:3.12"

​
Run The App

uvicorn app:app --reload --port 8080

​
Submit Code

curl -sS -X POST http://localhost:8080/run-python \
  -H "Content-Type: application/json" \
  -d '{"code":"print(sum([10, 20, 12]))","timeout":"10s"}'

​
Response

{
  "exit_code": 0,
  "stdout": "42\n",
  "stderr": "",
  "duration": "120ms"
}

​
Why This Pattern Works

  • The sandbox gets a short TTL so idle work is cleaned up.
  • The app destroys the sandbox in a finally block.
  • User code is written to a file instead of interpolated into a shell command.
  • Runtime failures are returned as normal execution results.
  • Provider failures return a clear API error.

​
Production Notes

Before exposing a code runner to users, add authentication to your app, per-user quotas, request size limits, audit logging, and runtime certification for the host you are using.