Skip to content

Startup and Shutdown

Startup order

  1. Verify or start the intended model provider.
  2. Set PIONEER_HOME in the backend terminal.
  3. Start the backend and wait for its ready message.
  4. Start the frontend.
  5. Open Pioneer and verify Backend Online.
  6. Open Machine and confirm profile, home, provider, and runtime evidence.

The frontend can stay running across an intentional backend restart and will reconnect. A browser refresh cannot load changed Python code into an old backend process.

Stop cleanly

  1. Press Ctrl+C in the frontend terminal.
  2. Press Ctrl+C in the backend terminal and wait for shutdown.
  3. Normally leave the separately managed Windows Ollama application running.

Durable state remains. Live streams, pending approvals, runtime-only mounts, resident Pioneer agents, and Machine samples do not survive backend shutdown. An independently running Ollama model may remain resident until its own expiry.