[codex] Support gpt-5.4 reasoning effort

[ShaoyeLi]

Summary

• use a supported OpenAI reasoning effort for openai/gpt-5.4 structured output calls
• keep existing defaults for Codex and earlier GPT-5 variants
• add a regression test that asserts the provider options passed to the AI SDK

Root cause

AISdkClient treated gpt-5.4 like older GPT-5 models and passed reasoningEffort: "minimal" to the OpenAI provider. gpt-5.4 rejects that value and only accepts none, low, medium, high, and xhigh, so structured output requests failed before the model call completed.

Testing

• HUSKY=0 corepack pnpm exec prettier --check lib/v3/llm/aisdk.ts tests/unit/aisdk-client-reasoning.test.ts
• HUSKY=0 corepack pnpm exec vitest run --config vitest.esm.config.mjs dist/esm/tests/unit/aisdk-client-reasoning.test.js

---

Summary by cubic

Fix structured output failures on openai/gpt-5.4 by using a supported reasoning effort. Defaults to reasoningEffort: "low" for this model, keeping existing behavior for Codex and earlier GPT-5 variants.

• Bug Fixes
  • Update model detection so openai/gpt-5.4 uses low reasoning effort (non-Codex), passing reasoningEffort: "low" and textVerbosity: "low" to OpenAI.
  • Add a regression test to assert provider options for openai/gpt-5.4 structured output calls.

^{Written for commit 5791761. Summary will update on new commits. Review in cubic}

[changeset-bot]

⚠️ No Changeset found

Latest commit: 5791761

Merging this PR will not cause a version bump for any packages. If these changes should not result in a new version, you're good to go. If these changes should result in a version bump, you need to add a changeset.

This PR includes no changesets

When changesets are added to this PR, you'll see the packages that this PR includes changesets for and the associated semver types

Click here to learn what changesets are, and how to add one.

Click here if you're a maintainer who wants to add a changeset to this PR

[github-actions]

This PR is from an external contributor and must be approved by a stagehand team member with write access before CI can run.
Approving the latest commit mirrors it into an internal PR owned by the approver.
If new commits are pushed later, the internal PR stays open but is marked stale until someone approves the latest external commit and refreshes it.

[ShaoyeLi]

Closing this duplicate exploratory PR in favor of #1920, which is the tracked contribution for the current workstream.