Unclear documentation on use_gemini_code_assist

[DinoChiesa]

TL;DR

This is a documentation bug. The docs state that there is a boolean flag use_gemini_code_assist. And the explanation is "when true, it says to use Gemini code Assist". What does that mean?

I think that "use VertexAI" implies that the Gemini CLI will Signin to a VertexAI project and use a URI like https://LOCATION-aiplatform.googleapis.com/v1/projects/PROJECT/locations/LOCATION/endpoints/ENDPOINT_ID:streamGenerateContent .

...whereas using just Gemini, I suppose it would be generativelanguage.googleapis.com ?

And then, if using code assist, it uses ... cloudcode-pa.googleapis.com ?

Expected behavior

No response

Observed behavior

No response

Action YAML

n/a

this is a documentation bug/question.

Log output

Additional information

No response

[gemini-cli]

🤖 Hi @DinoChiesa, I've received your request, and I'm working on it now! You can track my progress in the logs for more details.

[jerop]

@DinoChiesa yes, it means that it'd use code assist for inference as described in https://github.com/google-gemini/gemini-cli/blob/main/docs/cli/authentication.md#authentication-setup