Runtime Control and Security
The local runtime model, execution controls, read-only modes, and security boundaries in Raydo.
Raydo's local-first positioning only matters if the runtime is controllable. This is why runtime management and operating boundaries are part of the product itself.
Runtime and operating control
Raydo includes support for:
- OpenClaw local runtime management
- environment preflight and guided installation
- service state, ports, logs, and diagnostics
- portable execution and data isolation
Assistant control policies
Execution can be constrained through policies such as:
read-onlyconfirmexecute
Those policies can be paired with finer controls around file read, file write, terminal execution, and action confirmation.
Read-only protection and audit
- The Operations surface can expose system-wide read-only mode.
- Assistant actions can be reviewed through audit logs and exports.
- This makes powerful automation more governable and easier to trace back.
Web remote mode
Raydo also supports a web remote mode aimed at inspection and acceptance scenarios. It is intentionally read-only by default, so it does not blur into a full remote-write control panel.
Why it matters
Raydo's promise is not simply "more power." It is "more power with explicit boundaries." That is what makes local execution credible for real usage.