Quick Start
- Install Rio using installer scripts or pip/pipx.
- You will be redirected immediately to rio configure for mandatory onboarding.
- Choose your cloud/dashboard port when prompted (default is 8080).
- Start runtime with rio run (starts cloud + local).
- Open dashboard at http://localhost:<port>/dashboard.
Boot and Services
Linux setup can create user-level systemd services if you approve the start-on-boot prompt.
Generated services:
- rio-cloud.service
- rio-local.service
They are stored under ~/.config/systemd/user.
CLI Commands
- rio config get <path> - read config value
- rio config set <path> <value> - update config value
- rio config show - print entire config
- rio configure - mandatory setup flow
- rio doctor - diagnose and auto-fix issues
- rio doctor --no-fix - report only
- rio doctor --test-api - include Gemini API probe
- rio run - run cloud + local
- rio run --foreground - local foreground mode
- rio run --background - force detached mode
- rio run --skip-configure - bypass configure gate
- rio logs --component both --lines 80 - read logs
- rio logs --component local --follow - live log tail
- rio status --component both - process health
- rio stop --component both - stop runtime
- rio stop --component both --force - force kill process tree
Deployment Checklist
- Run rio doctor and confirm acknowledgement.
- Run rio run then validate with rio status and rio logs.
- Build package: python -m build (from Rio-Agent root).
- Validate metadata: twine check dist/*.
- Push to GitHub and trigger publish-pypi.yml from release/workflow dispatch.
Troubleshooting
If dashboard is not loading, confirm your selected port is available and cloud is running.
If local client does not connect, verify cloud_url in rio/config.yaml matches your chosen port.
If API calls fail, open rio/cloud/.env and check GEMINI_API_KEY.
Expected completion message: Rio is live at: http://localhost:<port>/dashboard