Does this project provide a non-API-Key method of obtaining authorization tokens to an LLM inference provider services?
It seems many of the AI company provided client applications are using a form of Open Id Connect (OAuth2 auth-code with PKCE 3-legged flow) to authenticate and obtain an identity token that is being used for authorization to the LLM inference service. For example, OpenAI Codex Desktop, one can login using an enterprise login in the application. The enterprise company doesn't generate any API-Keys for any of its employees.