Allow linkage to "Trusted Access for Cyber program"

While doing (non-security) patch review with copilot CLI and model Claude Opus
5, I received this error:

Execution failed: CAPIError: 422 This content was flagged for possible
cybersecurity risk. If this seem wrong, try rephrasing your request. To get
authorized for security work, join the Trusted Access for Cyber program:
https://chatgpt.com/cyber (Request ID: AAAE:E5BF3:76D4E1:87C02B:6A6D089A)

The request involved a multi-agent workflow, so Copilot may have routed some
part of it through an OpenAI model or service despite the selected session
model. In any case, the prescribed remedy does not appear usable: enrollment
at chatgpt.com/cyber is associated with my OpenAI identity, while this Copilot
request is associated with my GitHub identity.

Please provide a supported way for eligible Copilot users to obtain Trusted
Access for OpenAI-backed requests made through Copilot. This might involve a
Copilot-specific enrollment, linking an approved OpenAI identity, or another
GitHub-managed verification mechanism.

Commencez par reproduire l’erreur 422 signalée dans un workflow multi-agent Copilot CLI et par examiner la manière dont l’inscription à Trusted Access est actuellement associée aux identités. La tâche est considérée comme terminée lorsqu’un utilisateur Copilot approuvé dispose d’un moyen pris en charge et géré par GitHub d’obtenir Trusted Access pour les requêtes basées sur OpenAI.