fix(openai-adapters): keep reasoning deltas out of assistant content

[siewcapital]

Summary

• map reasoning-delta stream parts to delta.reasoning_content in the Vercel stream converter
• keep delta.content reserved for actual assistant text deltas
• add regression checks for both single-event conversion and mixed stream conversion

Why

reasoning-delta was previously being emitted as content, which mixed thinking output into assistant text for Vercel adapter streams.

Testing

• pnpm exec vitest run src/test/vercelStreamConverter.test.ts (from packages/openai-adapters)

Closes #11069

[github-actions]

Thank you for your submission, we really appreciate it. Like many open-source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution. You can sign the CLA by just posting a Pull Request Comment same as the below format.

---

I have read the CLA Document and I hereby sign the CLA

---

_{You can retrigger this bot by commenting recheck in this Pull Request. Posted by the CLA Assistant Lite bot.}

[cubic-dev-ai]

No issues found across 2 files

[RomneyDa]

Nice fix! note CLA requirement. Also non-blocking nitpick but avoid any type if possible.