feat(as_llm): add named OrcaRouter provider

[XiaoHuo888-hue]

Summary

Add a named OrcaRouter provider to the as_llm component registry, mirroring the existing OpenAI-compatible wrappers.

OrcaRouter is an OpenAI-compatible gateway (base URL https://api.orcarouter.ai/v1). Until now, ReMe users could only reach it by repurposing the openai backend with a custom LLM_BASE_URL and a non-obvious model value. This PR registers a first-class orcarouter backend so the provider is visible and selectable by name:

LLM_BACKEND=orcarouter
LLM_MODEL_NAME=orcarouter/auto
LLM_API_KEY=sk-orca-...

LLM_BASE_URL stays optional and defaults to https://api.orcarouter.ai/v1.

It also runs gateway-level, zero-trust security for AI agents on the same endpoint — screening every prompt/response and governing every tool call on a default-deny basis, with no application code changes.

Changes

• reme/components/as_llm/__init__.py
  • OrcaRouterCredential: OpenAICredential subclass with a default base_url of https://api.orcarouter.ai/v1 and an empty-string fallback, so the shared base_url: ${LLM_BASE_URL:-} config key works without forcing users to set LLM_BASE_URL.
  • OrcaRouterAsLLM: registered as backend orcarouter in the built-in component registry.
• README.md, README_ZH.md, docs/en/quick_start.md, docs/zh/quick_start.md, example.env: document the new backend.
• tests/unit/test_orcarouter_provider.py: registration, default, fallback, and custom-base-url coverage.

Verification

• Unit tests: pytest tests/unit → 1078 passed (2 pre-existing failures in test_codex_agent_wrapper.py reproduce on clean main; they are sandbox-environment issues with the Codex MCP stdio bridge, unrelated to this change).
• Pre-commit hooks (black, flake8, pylint, add-trailing-comma, pyroma) all pass on the changed files.
• Live L3 check: built OrcaRouterAsLLM with a real OrcaRouter key and hit https://api.orcarouter.ai/v1/chat/completions with model orcarouter/auto; the provider returned a valid completion.

Disclosure: I'm an engineer on the OrcaRouter team.

[CLAassistant]

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.

---

XiaoHuo888-hue seems not to be a GitHub user. You need a GitHub account to be able to sign the CLA. If you have already a GitHub account, please add the email address used for this commit to your account.
_{You have signed the CLA already but the status is still pending? Let us recheck it.}

[jinliyl]

Thank you for the contribution and for providing tests and documentation.

Could you clarify why a dedicated orcarouter backend is necessary here? Since OrcaRouter exposes an OpenAI-compatible API, it appears to work with the existing backend already:

LLM_BACKEND=openai
LLM_MODEL_NAME=orcarouter/auto
LLM_API_KEY=sk-orca-...
LLM_BASE_URL=https://api.orcarouter.ai/v1

Thanks again!