feat: add qwen provider

[oborges]

Summary

• add Alibaba Qwen provider using DashScope's OpenAI-compatible API
• register provider name qwen
• add llm-bridge[qwen] optional extra
• add offline fake-SDK tests and Qwen example
• document Qwen install, credentials, and endpoint behavior

Testing

• .venv/bin/ruff check .
• .venv/bin/black --check .
• .venv/bin/mypy src/llm_bridge
• .venv/bin/pytest

Live smoke test

Verified with qwen-plus using the international DashScope endpoint.

[github-actions]

All contributors have signed the CLA ✍️ ✅
_{Posted by the CLA Assistant Lite bot.}

[oborges]

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

[opensource-SantanderAI]

Thanks @oborges — the Qwen provider looks good and follows the same clean pattern as the DeepSeek one.

We just merged #13 (DeepSeek), which touched the same files this PR modifies (registry.py, providers/openai_sdk.py, tests/test_cloud_providers.py, tests/test_registry.py, pyproject.toml, README.md, CONTRIBUTING.md), so this branch now has merge conflicts against main (mergeable: CONFLICTING).

Could you please rebase onto the latest main and resolve the conflicts? In particular the provider registry entries and the optional-extras list in pyproject.toml now need to include both deepseek and qwen. Once it's rebased and CI is green we'll be happy to merge. Thanks!

[opensource-SantanderAI]

LGTM: Qwen provider (optional extra, OpenAI-compatible DashScope), tests offline incluidos, CI verde tras rebase sobre main.