docs: add Atomic Chat to Local LLMs guide

[yanalialiuk]

Summary

Documents Atomic Chat as another OpenAI-compatible local backend for OpenHands, alongside LM Studio, Ollama, vLLM, and SGLang.

Atomic Chat runs a desktop local inference stack and exposes a single HTTP API (default http://127.0.0.1:1337/v1). OpenHands already supports this pattern via advanced LLM settings; this PR makes discovery and copy/paste configuration explicit in Local LLMs.

Changes

• Mention Atomic Chat in the Advanced: Alternative LLM Backends intro.
• Add Create an OpenAI-Compatible Endpoint with Atomic Chat: install, enable local API, large context guidance, discover model id via GET /v1/models, Docker (host.docker.internal:1337/v1) vs same-host base URL, placeholder API key, troubleshooting.
• Extend Configure OpenHands (Alternative Backends) with Atomic Chat model string, port 1337, and API key note (grouped with Ollama for placeholder keys).

References

• Upstream project: https://github.com/AtomicBot-ai/Atomic-Chat
• Website: https://atomic.chat/

Notes for reviewers

• No new screenshots (Atomic Chat UI changes across versions); steps stay generic and link to upstream README.
• Anchors in the new section point to existing headings on the same page.