[dschenk]

Summary

A lazy subagent whose required tools need authentication currently reports "Unable to initialize the selected subagent", even when its host knows that sign-in is required. Add two allowlisted host classifications: authentication_required and tools_unavailable. The former tells the parent to wait for sign-in before retrying; unknown classifications still receive the generic fallback.

This is the SDK part of a LibreChat integration fix. It does not infer causes from provider error text or change authentication, retry, or execution policy. The existing failure hook selects a category; all outward messages remain fixed strings.

Testing

• Negative control: both new classification cases fail on unchanged upstream (2 failed, 18 passed).
• Diagnostics and lazy-subagent suites: 72 passed.
• npx tsc --noEmit: passed.
• Touched-file ESLint and formatting: passed.
• The parameterized regressions also verify that error messages, names, stacks and token-like strings cannot enter the public diagnostic or retained resolution error.

No live provider or OAuth server was used. No UI changes in this library.

Host integration: LibreChat-AI/LibreChat#17302 (draft pending an SDK release). Reconnect UI: LibreChat-AI/LibreChat#17300. Tracking: LibreChat-AI/LibreChat#17301.

[dschenk]

@codex review the current head e570ae7, especially allowlisted diagnostics and preservation of the raw-error disclosure boundary. The LibreChat host integration will be linked separately.

[chatgpt-codex-connector]

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-10-21T09:19:26.712435Z	30a6657	Manual 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" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: e570ae75c6

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".

[dschenk]

@codex review the current head 30a6657. The documentation finding is addressed; focused diagnostics/lazy-executor tests (72) and TypeScript pass.

[chatgpt-codex-connector]

Codex Review: Didn't find any major issues. Chef's kiss.

Reviewed commit: 30a66572f1

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".