Add CLIENT SETNAME to Valkey client for connection identification

Nobody has claimed this yet.

## Summary

The llmware Valkey embeddings integration creates a `GlideClient` without setting a `client_name`. This means connections appear anonymous in monitoring tools like `CLIENT LIST` and Valkey Admin.

## Suggested Fix

In [`tests/embeddings/test_valkey_embeddings.py` (line ~37)](https://github.com/llmware-ai/llmware/blob/3b848a4d2d5c11c0071ee413add4a04276d9112d/tests/embeddings/test_valkey_embeddings.py#L37), add `client_name` to the `GlideClientConfiguration`:

```python
cfg = GlideClientConfiguration(
[NodeAddress(host, port)],
client_name="llmware_embedding_client",
)
```

## Why This Matters

When monitoring a Valkey server with multiple connected applications, `CLIENT LIST` shows each connection's name. Without a client name, operators cannot distinguish llmware connections from other anonymous clients. This is especially important in production environments with ElastiCache where multiple services share the same Valkey cluster.

Setting `client_name` sends a `CLIENT SETNAME` command on connection, making the connection identifiable in:
- `CLIENT LIST` output
- Monitoring dashboards (e.g., Valkey Admin)
- CloudWatch metrics (ElastiCache)

## Naming Convention

Suggested client name: `llmware_embedding_client`

Pattern: `{project}_{purpose}_client`
- `llmware` = project name
- `embedding` = purpose (embedding storage and retrieval)
- `_client` = convention (matches Valkey Admin naming pattern)

Start in tests/embeddings/test_valkey_embeddings.py around the GlideClientConfiguration setup and review how the Valkey embedding client is configured. Run the relevant embeddings test, then verify the connection uses the requested llmware_embedding_client name in Valkey monitoring output.

Written by the indexing model from the issue text.