First launch checklist
- · Sign-in authenticates you with the product; integrations still need OAuth one by one.
- · Most people can ignore the Advanced panel unless they self-host infrastructure.
- · Expect macOS Accessibility, Input Monitoring, and optional camera/microphone prompts.
1 · Open the app deliberately
Early beta builds change quickly. If Gatekeeper or Defender complains, pause and confirm the binary matches the official distribution flow (see README in https://github.com/tinyhumansai/openhuman).
2 · OS permissions
- macOS requests Accessibility, Input Monitoring for the voice hotkey, and Camera/Microphone when Meeting Agent / voice enters the picture.
- Windows concentrates on Defender/SmartScreen at install time but may still surface microphone or camera prompts when those features activate.
- Revisit allowances under Settings → Automation & Channels anytime.
3 · Sign-in is not Gmail access yet
The Getting Started docs describe the splash copy as roughly Sign in — let's cook. That step wires you into the product accounts and subscription surfaces. Third-party scopes still require explicit OAuth approvals per connector.
4 · Advanced core RPC URL
Power users running private infrastructure can redirect the desktop app to a custom RPC endpoint inside the Advanced panel. If you did not spin up your own backend, leave this untouched—misconfigured URLs surface as endless reconnect loops during onboarding.
5 · Proceed to connectors
Continue with Integrations & sync next so you understand the ~20 minute auto-fetch rhythm before typing your first briefing prompt.