feat: add Phase F.1 memory infrastructure

Add multi-tenancy support and memory storage infrastructure:

- Add ContextVar-based request context (src/core/context.py)
  - Async-safe user/conversation tracking via contextvars
  - RequestContext manager for clean setup/teardown
  - get_user(), get_conversation_id() helpers

- Add multi-tenancy utilities (src/core/multi_tenancy.py)
  - User ID sanitization for collection/key names
  - get_memory_collection_name(), get_session_key() helpers

- Add Ollama embedding client (src/core/embeddings.py)
  - nomic-embed-text model (768 dimensions)
  - embed(), embed_batch(), health_check() methods

- Add Qdrant client wrapper (src/core/qdrant.py)
  - Per-user collection pattern: memories_{user}
  - upsert_memory(), search_memories(), delete_memory()
  - Type-based filtering support

- Add Redis memory cache (src/core/memory_cache.py)
  - Session context with 24h TTL
  - Recent entities tracking
  - Separate from benchmarks (db=2)

- Update config with memory settings
  - QDRANT_HOST, QDRANT_PORT, QDRANT_EMBEDDING_DIM
  - OLLAMA_EMBEDDING_MODEL
  - REDIS_MEMORY_DB, REDIS_MEMORY_TTL_HOURS

- Add user field to ResponseRequest (OpenAI standard)
- Set context in router, reset in finally block
- Update librarian client to use get_user() (12 methods)

All 333 unit tests pass.

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

src/responses/router.py

@@ -11,6 +11,7 @@ from sse_starlette.sse import EventSourceResponse
 from src.responses import service
 from src.responses.schemas import ResponseRequest, Response
 from src.core.exceptions import ModelNotFoundError, AppException
+from src.core.context import current_user, current_conversation
 
 logger = logging.getLogger(__name__)
 
@@ -94,6 +95,11 @@ async def create_response(
     """
     logger.info(f"Response request for model: {request.model}")
 
+    # Set request context (propagates through all async calls)
+    user_token = current_user.set(request.user or "jpmschweitzer")
+    conv_id = request.metadata.get("conversation_id") if request.metadata else None
+    conv_token = current_conversation.set(conv_id)
+
     try:
         # Check if this is a Tatlock request - use Steward preprocessing (Phase 2)
         model_id = request.model
@@ -136,3 +142,8 @@ async def create_response(
     except Exception as e:
         logger.error(f"Unexpected error: {e}", exc_info=True)
         raise HTTPException(status_code=500, detail="Internal server error")
+
+    finally:
+        # Reset context (important for connection reuse)
+        current_user.reset(user_token)
+        current_conversation.reset(conv_token)

src/responses/schemas.py

@@ -138,6 +138,10 @@ class ResponseRequest(CustomBaseModel):
         default=None,
         description="Stop sequences"
     )
+    user: str | None = Field(
+        default=None,
+        description="Unique identifier for end-user (OpenAI standard)"
+    )
 
     @field_validator('reasoning')
     @classmethod