Add remote OpenAI-compatible OpenChat example (DaoXE)

[seven7763]

Summary

• Adds solutions/models/using-daoxe-remote.py, a self-contained example that registers a remote OpenAI-compatible Chat Completions endpoint via existing ModelCatalog().register_open_chat_model(..., api_base=...) (same path as LM Studio / local OpenChat).
• Extends docs/components/model_catalog.md with a short “remote OpenAI-compatible endpoint” section pointing at that example.
• Uses env vars only: DAOXE_API_KEY, DAOXE_MODEL, optional DAOXE_BASE_URL (default https://daoxe.com/v1). No hardcoded model IDs (catalog / account-dependent).

DaoXE is a multi-model, multi-protocol API gateway (OpenAI Chat Completions / Responses and Anthropic Messages / Claude protocol, plus other catalog endpoints). This PR only documents the OpenAI-compatible surface that llmware’s OpenChatModel already supports—no core library changes.

Why this fits

llmware already documents local OpenAI-compatible servers (LM Studio) via register_open_chat_model. Remote gateways use the same API; this example makes that explicit for contributors who want cloud multi-model access without changing llmware internals.

How to try

export DAOXE_API_KEY="your_api_key"
export DAOXE_MODEL="model_id_from_your_live_catalog"
# optional: export DAOXE_BASE_URL="https://daoxe.com/v1"
python solutions/models/using-daoxe-remote.py

Notes

• Service availability can vary by account and region; DaoXE does not serve users in mainland China.
• Example exits cleanly with setup instructions if env vars are missing (safe for CI / dry runs without credentials).

Test plan

• python3 -m py_compile solutions/models/using-daoxe-remote.py
• Diff limited to docs + one new example under solutions/models/
• Maintainer: optional live run with a DaoXE key against a catalog model ID

References

• https://daoxe.com
• https://daoxe.com/pricing
• Existing local path: solutions/models/using-open-chat-models.py