Skip to main content
This page gets the opensre command onto your laptop or server, then walks you through the short setup so you can run investigations.

What you get

After install, you have:
  • The opensre CLI on your PATH
  • An interactive shell when you run opensre with no arguments
  • An onboard wizard to pick your LLM and connect tools like Datadog or Slack
No sudo required on macOS/Linux in the usual case — the installer puts the binary somewhere writable (often ~/.local/bin) and tells you if you need a PATH tweak.

Install

Pick one method. One-line installer (most people):
In a normal interactive terminal, the installer often starts opensre onboard for you when it finishes. In CI or scripts it skips that and prints what to do next. Don’t want auto-onboard?
Homebrew:
More options (PATH notes, skipping GitHub CLI install, on-prem layout): Install locally.

Set things up

If onboard didn’t start automatically, run it yourself:
The wizard asks for:
  1. An LLM — OpenAI, Anthropic, a local model, or another provider you already use
  2. Your tools — whatever you want investigations to query (Datadog, Grafana, Slack, and so on)
  3. Optional extras — masking and messaging, if you want them
Then check that the tools you connected actually work:

Try a quick investigation

Or open the shell and describe what’s going on:
Full walkthrough with GIFs: Quickstart.

Environment-specific guides

Stuck?

  • Command not found — open a new terminal, or add the bin directory the installer printed to your PATH
  • Onboard can’t reach your LLM — check the key and network; you can save and retry later
  • Docker / make missing — see Quickstart troubleshooting
Need deploy options beyond a laptop binary? See Deployment.