Deny first
If a host is not approved, the flow should stop and explain what is missing. Silent cross-domain control is not the model.
RTO treats allow-listing as part of the product contract, not as an afterthought. A page cannot drive arbitrary domains by default: trusted hosts are declared explicitly, and that trust boundary stays visible in the extension itself.
If a host is not approved, the flow should stop and explain what is missing. Silent cross-domain control is not the model.
Trusted hosts are declared by the user. This keeps the control surface understandable, reviewable and bounded.
RTO works with visible tabs and visible safety signals, so the user can see what is being controlled.
Start with the minimal demo and observe the deny-by-default behavior if the target host has not been allowed yet.
Control is explicit. If a host is not allowed, the flow should stop, explain the missing approval, and retry only after the user authorizes it.
RTO does not ask for vague trust. It exposes a clear rule set where allowed hosts are explicit and everything else stays outside the control surface.
RTO works best when you want real visible tabs, local orchestration, and explicit allow-list control rather than background relay infrastructure.
Only explicitly allowed hosts can be controlled. That makes the product easier to explain and easier to approve internally.
The extension remains the visible control point for allowed domains, localhost policies and related safeguards.
Bounded browser control is a commercial advantage when users are cautious about automation plugins.
Start fast with the browser-side SDK.
Open SDKRead the public contract, allow-list notes and examples.
Open docsTry copy-paste starter examples and live demos.
Browse examplesSee where RTO fits against recorder tools and Selenium-style stacks.
Open compare