Skip to main content
This tutorial gets you from “installed” to a real execution showing up in the dashboard.

​
Prerequisites

​
Step 1: Install

pip install omium

​
Step 2: Authenticate

If you’re on a laptop, the fastest setup is: 
omium init
If you prefer environment variables: 
export OMIUM_API_KEY=omium_your_key_here
export OMIUM_API_URL=https://api.omium.ai

​
Step 3: Run a minimal traced script

Create a file named quickstart.py: 
import omium

omium.init()  # reads ~/.omium/config.json or OMIUM_API_KEY/OMIUM_API_URL

@omium.trace("hello")
def hello(name: str) -> str:
    return f"hello, {name}"

if __name__ == "__main__":
    print(hello("omium"))
Run it: 
python quickstart.py
Expected output: 
hello, omium

​
Step 4: See it in the dashboard

Open app.omium.ai and look for the execution under:
  • Executions
  • or your project area in Automations (once you start pushing projects)
If you don’t see it after a minute:
  • confirm your key is valid (omium init again)
  • confirm your API URL is https://api.omium.ai
  • check network access from your machine

​
What’s next

LangGraph

Auto-instrument LangGraph calls like invoke() and stream().

CrewAI

Trace kickoff() and multi-agent runs.

First project

Push a project so it appears as an Automation in the app.

CLI reference

Inspect executions, stream logs, replay, and more.