Security Warning Dialog Appears When Using window.openai.openExternal

[yanbai]

Issue Description
I'm encountering a "Check this link is safe" security warning dialog every time I use [window.openai.openExternal] to open third-party pages in my ChatGPT widget. While I understand the security reasoning behind this warning, it creates friction in the user experience for legitimate use cases.

Problem
The security warning dialog appears for every external link, even when:

The destination is a trusted, well-known domain
The widget is designed specifically for navigation to external services
The links are part of the core functionality, not unexpected redirects

Request
While security is important, could the validation be more nuanced to balance security with user experience? Perhaps allow trusted domains with appropriate warnings, or validate based on the widget's actual functionality rather than applying blanket restrictions to all external navigation?

The goal is to maintain security while enabling smooth user flows for legitimate business use cases.