Introduce ModelsManager and migrate app-server to use it.
#7552
Conversation
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| pub async fn list_models(&self) -> Vec<ModelPreset> { | ||
| self.models_manager.available_models.clone() | ||
| } |
There was a problem hiding this comment.
ConversationManager::list_models() simply clones models_manager.available_models, which is built once in ConversationManager::new using the auth mode present at startup. There is no refresh when the auth state changes, so if the process starts unauthenticated and the user later logs in with an API key (or switches auth modes), ModelList responses continue to return the initial preset set (e.g., still exposing gpt-5.1-codex-max to API-key users) until restart. The model list should be computed from the current auth mode instead of a startup snapshot.
Useful? React with 👍 / 👎.
|
|
||
| impl ModelsManager { | ||
| pub fn new(auth_mode: Option<AuthMode>) -> Self { | ||
| std::thread::sleep(Duration::from_secs(1)); |
There was a problem hiding this comment.
I think there should be an async method load models and we should async sleep there
There was a problem hiding this comment.
Or remove completely for now
No description provided.