fix: override verbosity for gpt-5-codex #6007
Conversation
![chatgpt-codex-connector[bot]](https://cdn.statically.io/img/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".
|
@codex review this |
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".
| base_instructions: GPT_5_CODEX_INSTRUCTIONS.to_string(), | ||
| apply_patch_tool_type: Some(ApplyPatchToolType::Freeform), | ||
| support_verbosity: true, | ||
| support_verbosity: false, |
There was a problem hiding this comment.
Setting support_verbosity to false for gpt-5-codex routes every call into the else branch in core/src/client.rs (lines 216-223). That branch unconditionally emits a warn! about model_verbosity being set, even when self.config.model_verbosity is None (the default for most users). As a result, every request against gpt-5-codex will now log a misleading warning. This regression was not present before this change (the branch was never taken), so we should either keep support_verbosity true and clamp values, or make the warning conditional on model_verbosity actually being set.
Useful? React with 👍 / 👎.
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".
| base_instructions: GPT_5_CODEX_INSTRUCTIONS.to_string(), | ||
| apply_patch_tool_type: Some(ApplyPatchToolType::Freeform), | ||
| support_verbosity: true, | ||
| support_verbosity: false, |
There was a problem hiding this comment.
Turning support_verbosity off for gpt-5-codex causes stream_responses to take the warning branch in core/src/client.rs line 216 on every call. Because the default Config leaves model_verbosity as None, we now emit warn!("model_verbosity is set but ignored…") even when the user never configured it. gpt-5-codex is our default model, so this introduces a new, user-visible warning on every request. This regression didn’t exist before flipping the flag; we either need to keep support_verbosity true or update the warning path to only fire when the option is actually set.
Useful? React with 👍 / 👎.
|
LGTM, but you might want to look at the issues that codex raised in its code review. |
we are seeing reports of users having verbosity in their config.toml and facing issues. gpt-5-codex doesn't accept other values rather than medium for verbosity.