phase 3 adiitons for qdrant support

Author: jaschadub

pyproject.toml

@@ -37,6 +37,7 @@ dependencies = [
     "pyyaml>=6.0.0",
     "pydantic-settings>=2.0.0",
     "redis>=5.0.0",
+    "qdrant-client>=1.7.0",
 ]
 
 [project.optional-dependencies]

requirements.txt

@@ -5,4 +5,5 @@ pyjwt>=2.8.0
 cryptography>=41.0.0
 redis>=5.0.0
 pyyaml>=6.0.0
-pydantic-settings>=2.0.0
+pydantic-settings>=2.0.0
+qdrant-client>=1.7.0

symbiont/client.py

@@ -26,6 +26,10 @@
     AgentMetrics,
     AgentStatusResponse,
     AnalysisResults,
+    # Phase 3 Qdrant Integration models
+    CollectionCreateRequest,
+    CollectionInfo,
+    CollectionResponse,
     ConsolidationResponse,
     # Configuration models (Phase 1)
     ContextQuery,
@@ -66,8 +70,10 @@
     SigningRequest,
     SigningResponse,
     SystemMetrics,
+    UpsertResponse,
     VectorSearchRequest,
     VectorSearchResponse,
+    VectorUpsertRequest,
     WebhookTriggerRequest,
     WebhookTriggerResponse,
     WorkflowExecutionRequest,
@@ -1147,3 +1153,131 @@ def list_agent_memories(self, agent_id: str, limit: int = 100) -> MemorySearchRe
         }
         response = self._request("GET", "memory/agent", params=params)
         return MemorySearchResponse(**response.json())
+
+    # =============================================================================
+    # Phase 3 Qdrant Vector Database Methods
+    # =============================================================================
+
+    def create_vector_collection(self, collection_request: Union[CollectionCreateRequest, Dict[str, Any]]) -> CollectionResponse:
+        """Create a new vector collection.
+
+        Args:
+            collection_request: Collection creation request
+
+        Returns:
+            CollectionResponse: Collection creation result
+        """
+        if isinstance(collection_request, dict):
+            collection_request = CollectionCreateRequest(**collection_request)
+
+        response = self._request("POST", "vectors/collections", json=collection_request.model_dump())
+        return CollectionResponse(**response.json())
+
+    def delete_vector_collection(self, collection_name: str) -> Dict[str, Any]:
+        """Delete a vector collection.
+
+        Args:
+            collection_name: Name of the collection to delete
+
+        Returns:
+            Dict[str, Any]: Deletion confirmation
+        """
+        response = self._request("DELETE", f"vectors/collections/{collection_name}")
+        return response.json()
+
+    def get_collection_info(self, collection_name: str) -> CollectionInfo:
+        """Get information about a vector collection.
+
+        Args:
+            collection_name: Name of the collection
+
+        Returns:
+            CollectionInfo: Collection information
+        """
+        response = self._request("GET", f"vectors/collections/{collection_name}")
+        return CollectionInfo(**response.json())
+
+    def list_vector_collections(self) -> List[str]:
+        """List all vector collections.
+
+        Returns:
+            List[str]: List of collection names
+        """
+        response = self._request("GET", "vectors/collections")
+        return response.json()
+
+    def add_vectors(self, upsert_request: Union[VectorUpsertRequest, Dict[str, Any]]) -> UpsertResponse:
+        """Add vectors to a collection.
+
+        Args:
+            upsert_request: Vector upsert request
+
+        Returns:
+            UpsertResponse: Upsert operation result
+        """
+        if isinstance(upsert_request, dict):
+            upsert_request = VectorUpsertRequest(**upsert_request)
+
+        response = self._request("POST", "vectors/upsert", json=upsert_request.model_dump())
+        return UpsertResponse(**response.json())
+
+    def get_vectors(self, collection_name: str, vector_ids: List[Union[str, int]]) -> List[Dict[str, Any]]:
+        """Get vectors by IDs from a collection.
+
+        Args:
+            collection_name: Name of the collection
+            vector_ids: List of vector IDs to retrieve
+
+        Returns:
+            List[Dict[str, Any]]: Retrieved vectors
+        """
+        params = {
+            "collection_name": collection_name,
+            "ids": vector_ids
+        }
+        response = self._request("GET", "vectors/retrieve", params=params)
+        return response.json()
+
+    def search_vectors(self, search_request: Union[VectorSearchRequest, Dict[str, Any]]) -> VectorSearchResponse:
+        """Search vectors using similarity search.
+
+        Args:
+            search_request: Vector search request
+
+        Returns:
+            VectorSearchResponse: Search results
+        """
+        if isinstance(search_request, dict):
+            search_request = VectorSearchRequest(**search_request)
+
+        response = self._request("POST", "vectors/search", json=search_request.model_dump())
+        return VectorSearchResponse(**response.json())
+
+    def delete_vectors(self, collection_name: str, vector_ids: List[Union[str, int]]) -> Dict[str, Any]:
+        """Delete vectors from a collection.
+
+        Args:
+            collection_name: Name of the collection
+            vector_ids: List of vector IDs to delete
+
+        Returns:
+            Dict[str, Any]: Deletion confirmation
+        """
+        data = {
+            "collection_name": collection_name,
+            "ids": vector_ids
+        }
+        response = self._request("DELETE", "vectors/delete", json=data)
+        return response.json()
+
+    def count_vectors(self, collection_name: str) -> int:
+        """Count vectors in a collection.
+
+        Args:
+            collection_name: Name of the collection
+
+        Returns:
+            int: Number of vectors in the collection
+        """
+        response = self._request("GET", f"vectors/collections/{collection_name}/count")
+        return response.json()["count"]

symbiont/models.py

@@ -2,7 +2,7 @@
 
 from datetime import datetime
 from enum import Enum
-from typing import Any, Dict, List, Optional
+from typing import Any, Dict, List, Optional, Union
 
 from pydantic import BaseModel, Field
 
@@ -788,3 +788,99 @@ class MemorySearchResult(BaseModel):
     relevance_score: float = Field(..., description="Relevance score for the search query")
     match_reason: str = Field(..., description="Reason for the match")
     highlighted_content: Optional[Dict[str, Any]] = Field(None, description="Content with search highlights")
+
+
+# =============================================================================
+# Phase 3 Qdrant Integration Models
+# =============================================================================
+
+class Vector(BaseModel):
+    """Vector representation for Qdrant."""
+    id: Union[str, int] = Field(..., description="Vector identifier")
+    values: List[float] = Field(..., description="Vector values/embeddings")
+    metadata: Dict[str, Any] = Field(default_factory=dict, description="Vector metadata")
+
+
+class Point(BaseModel):
+    """Point representation for Qdrant."""
+    id: Union[str, int] = Field(..., description="Point identifier")
+    vector: Union[List[float], Dict[str, List[float]]] = Field(..., description="Vector data")
+    payload: Dict[str, Any] = Field(default_factory=dict, description="Point payload/metadata")
+
+
+class SearchQuery(BaseModel):
+    """Search query for vector similarity search."""
+    vector: List[float] = Field(..., description="Query vector")
+    limit: int = Field(10, description="Maximum number of results")
+    score_threshold: Optional[float] = Field(None, description="Minimum similarity score")
+    filter: Optional[Dict[str, Any]] = Field(None, description="Payload filter conditions")
+    with_payload: bool = Field(True, description="Include payload in results")
+    with_vector: bool = Field(False, description="Include vectors in results")
+
+
+class CollectionCreateRequest(BaseModel):
+    """Request to create a new vector collection."""
+    name: str = Field(..., description="Collection name")
+    vector_size: int = Field(..., description="Vector dimension size")
+    distance: str = Field("Cosine", description="Distance metric (Cosine, Euclidean, Dot)")
+    on_disk_payload: bool = Field(False, description="Store payload on disk")
+    hnsw_config: Optional[Dict[str, Any]] = Field(None, description="HNSW configuration")
+    optimizers_config: Optional[Dict[str, Any]] = Field(None, description="Optimizer configuration")
+
+
+class CollectionResponse(BaseModel):
+    """Response from collection operations."""
+    collection_name: str = Field(..., description="Collection name")
+    status: str = Field(..., description="Operation status")
+    result: Optional[Dict[str, Any]] = Field(None, description="Operation result details")
+
+
+class CollectionInfo(BaseModel):
+    """Information about a vector collection."""
+    collection_name: str = Field(..., description="Collection name")
+    config: Dict[str, Any] = Field(..., description="Collection configuration")
+    status: str = Field(..., description="Collection status")
+    vectors_count: int = Field(..., description="Number of vectors")
+    indexed_vectors_count: int = Field(..., description="Number of indexed vectors")
+    points_count: int = Field(..., description="Number of points")
+
+
+class VectorUpsertRequest(BaseModel):
+    """Request to upsert vectors into a collection."""
+    collection_name: str = Field(..., description="Target collection name")
+    points: List[Point] = Field(..., description="Points to upsert")
+    wait: bool = Field(True, description="Wait for operation completion")
+
+
+class UpsertResponse(BaseModel):
+    """Response from vector upsert operation."""
+    collection_name: str = Field(..., description="Collection name")
+    operation_id: Optional[str] = Field(None, description="Operation identifier")
+    status: str = Field(..., description="Operation status")
+    points_count: int = Field(..., description="Number of points processed")
+
+
+class VectorPoint(BaseModel):
+    """Vector point with metadata."""
+    id: Union[str, int] = Field(..., description="Point identifier")
+    vector: List[float] = Field(..., description="Vector values")
+    payload: Dict[str, Any] = Field(default_factory=dict, description="Point metadata")
+    score: Optional[float] = Field(None, description="Similarity score (for search results)")
+
+
+class EmbeddingRequest(BaseModel):
+    """Request to generate embeddings."""
+    texts: List[str] = Field(..., description="Texts to embed")
+    model: str = Field("default", description="Embedding model to use")
+    collection_name: Optional[str] = Field(None, description="Target collection")
+    metadata: Optional[List[Dict[str, Any]]] = Field(None, description="Metadata for each text")
+
+
+class EmbeddingResponse(BaseModel):
+    """Response from embedding generation."""
+    embeddings: List[List[float]] = Field(..., description="Generated embeddings")
+    model: str = Field(..., description="Model used for embedding")
+    processing_time_ms: float = Field(..., description="Processing time in milliseconds")
+    token_count: Optional[int] = Field(None, description="Total token count processed")
+
+