### Check for existing issues

- [x] I have searched the existing issues and checked that my issue is not a duplicate.

### What happened?

When routing requests through LiteLLM to Anthropic models, the Anthropic API returns a 400 error if the request body contains `vector_store_ids` or `vector_store_id`:

``` litellm.BadRequestError: AnthropicException - b'{"type":"error","error":{"type":"invalid_request_error","message":"vector_store_ids: Extra inputs are not permitted"}}' ```

**Root cause**: LiteLLM proxy injects `vector_store_ids` into the request body for internal RBAC / vector-store pre-call hooks (see `proxy/auth/auth_checks.py` and `proxy/common_utils/http_parsing_utils.py`). This parameter is intentionally an OpenAI Assistants API concept that Anthropic does not recognise.

In `AnthropicConfig.transform_request()` the final payload is built with:

```python data = { "model": model, "messages": anthropic_messages, **optional_params, # ← vector_store_ids leaks through here } ```

`is_vertex_request` is already stripped before this point, but `vector_store_ids` / `vector_store_id` are not.

### Steps to Reproduce

```python import litellm

response = litellm.…