You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository was archived by the owner on Sep 28, 2025. It is now read-only.
GitHub Copilot makes GPT-5 available on pro plan. now to support the model in tandem, you got to edit the copilot.go which is like adding a line.
Google will release Gemini 3 model series, will discontinue some of the older Gemini series like 1.5, 2.5, 2.0 and then you got to edit it accordingly.
so i want to automate this. how something like this could be done, i don't know honestly.
what i think we can do here is add the HTTP response into the prompt after querying the corresponding endpoint from these LLM providers to query active model ids. this doesn't necessarily have to be github action like gemini-cli though. we use it as a prompt template or some in IDE. whatever works in predictable and fast in terms of iteration.
take couple of scenarios first.
so i want to automate this. how something like this could be done, i don't know honestly.
what i think we can do here is add the HTTP response into the prompt after querying the corresponding endpoint from these LLM providers to query active model ids. this doesn't necessarily have to be github action like gemini-cli though. we use it as a prompt template or some in IDE. whatever works in predictable and fast in terms of iteration.