feat(extension): add OrcaRouter as a named dense embedding provider

[XiaoHuo888-hue]

Summary

Add OrcaRouter as a named dense-embedding provider in the Python extension layer, mirroring the existing OpenAIDenseEmbedding / QwenDenseEmbedding provider pattern.

OrcaRouter is an OpenAI-compatible gateway that exposes a broad catalog of frontier models through a single endpoint (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

• python/zvec/extension/orcarouter_function.py: new OrcaRouterFunctionBase base class with DEFAULT_ORCAROUTER_BASE_URL, ORCAROUTER_API_KEY env-var handling, and embedding API call/response parsing. Reuses the official openai client (no new dependency).
• python/zvec/extension/orcarouter_embedding_function.py: new OrcaRouterDenseEmbedding class with a model dimension table and LRU caching.
• python/zvec/extension/__init__.py and python/zvec/__init__.py: export the new classes.
• python/tests/test_embedding.py: add TestOrcaRouterDenseEmbedding mirroring the OpenAI test suite (init, env key, empty input, success, API error, invalid response, dimension mismatch, callable, custom base URL, real-API integration test).

Usage

from zvec.extension import OrcaRouterDenseEmbedding
import os

os.environ["ORCAROUTER_API_KEY"] = "sk-orca-..."
emb = OrcaRouterDenseEmbedding(model="openai/text-embedding-3-small")
vector = emb.embed("Hello, world!")

Validation

• ruff check and ruff format --check pass on all changed files.
• pytest python/tests/test_embedding.py -k OrcaRouter: 17 passed, 1 skipped (integration test gated behind ZVEC_RUN_INTEGRATION_TESTS=1).
• Existing embedding tests unaffected: full test_embedding.py run — 94 passed, 13 skipped (pre-existing integration tests).
• L3 live test: ran OrcaRouterDenseEmbedding with a real key against https://api.orcarouter.ai/v1/embeddings (model openai/text-embedding-3-small) → HTTP 200, 1536-dim vector returned.

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.}