agno 2.1.4__py3-none-any.whl → 2.1.5__py3-none-any.whl

agno/tools/googlesheets.py

@@ -48,13 +48,14 @@ import json
 from functools import wraps
 from os import getenv
 from pathlib import Path
-from typing import Any, List, Optional
+from typing import Any, List, Optional, Union
 
 from agno.tools import Toolkit
 
 try:
     from google.auth.transport.requests import Request
     from google.oauth2.credentials import Credentials
+    from google.oauth2.service_account import Credentials as ServiceAccountCredentials
     from google_auth_oauthlib.flow import InstalledAppFlow
     from googleapiclient.discovery import Resource, build
 except ImportError:
@@ -91,9 +92,10 @@ class GoogleSheetsTools(Toolkit):
         scopes: Optional[List[str]] = None,
         spreadsheet_id: Optional[str] = None,
         spreadsheet_range: Optional[str] = None,
-        creds: Optional[Credentials] = None,
+        creds: Optional[Union[Credentials, ServiceAccountCredentials]] = None,
         creds_path: Optional[str] = None,
         token_path: Optional[str] = None,
+        service_account_path: Optional[str] = None,
         oauth_port: int = 0,
         enable_read_sheet: bool = True,
         enable_create_sheet: bool = False,
@@ -108,9 +110,10 @@ class GoogleSheetsTools(Toolkit):
             scopes (Optional[List[str]]): Custom OAuth scopes. If None, uses write scope by default.
             spreadsheet_id (Optional[str]): ID of the target spreadsheet.
             spreadsheet_range (Optional[str]): Range within the spreadsheet.
-            creds (Optional[Credentials]): Pre-existing credentials.
+            creds (Optional[Credentials | ServiceAccountCredentials]): Pre-existing credentials.
             creds_path (Optional[str]): Path to credentials file.
             token_path (Optional[str]): Path to token file.
+            service_account_path (Optional[str]): Path to a service account file.
             oauth_port (int): Port to use for OAuth authentication. Defaults to 0.
             enable_read_sheet (bool): Enable reading from a sheet.
             enable_create_sheet (bool): Enable creating a sheet.
@@ -126,6 +129,7 @@ class GoogleSheetsTools(Toolkit):
         self.token_path = token_path
         self.oauth_port = oauth_port
         self.service: Optional[Resource] = None
+        self.service_account_path = service_account_path
 
         # Determine required scopes based on operations if no custom scopes provided
         if scopes is None:
@@ -171,6 +175,17 @@ class GoogleSheetsTools(Toolkit):
         if self.creds and self.creds.valid:
             return
 
+        service_account_path = self.service_account_path or getenv("GOOGLE_SERVICE_ACCOUNT_FILE")
+
+        if service_account_path:
+            self.creds = ServiceAccountCredentials.from_service_account_file(
+                service_account_path,
+                scopes=self.scopes,
+            )
+            if self.creds and self.creds.expired:
+                self.creds.refresh(Request())
+            return
+
         token_file = Path(self.token_path or "token.json")
         creds_file = Path(self.credentials_path or "credentials.json")
 
@@ -178,7 +193,7 @@ class GoogleSheetsTools(Toolkit):
             self.creds = Credentials.from_authorized_user_file(str(token_file), self.scopes)
 
         if not self.creds or not self.creds.valid:
-            if self.creds and self.creds.expired and self.creds.refresh_token:
+            if self.creds and self.creds.expired and self.creds.refresh_token:  # type: ignore
                 self.creds.refresh(Request())
             else:
                 client_config = {
@@ -199,7 +214,7 @@ class GoogleSheetsTools(Toolkit):
                     flow = InstalledAppFlow.from_client_config(client_config, self.scopes)
                 # Opens up a browser window for OAuth authentication
                 self.creds = flow.run_local_server(port=self.oauth_port)
-            token_file.write_text(self.creds.to_json()) if self.creds else None
+            token_file.write_text(self.creds.to_json()) if self.creds else None  # type: ignore
 
     @authenticate
     def read_sheet(self, spreadsheet_id: Optional[str] = None, spreadsheet_range: Optional[str] = None) -> str:

agno/tools/mcp_toolbox.py

@@ -35,6 +35,7 @@ class MCPToolbox(MCPTools, metaclass=MCPToolsMeta):
         tool_name: Optional[str] = None,
         headers: Optional[Dict[str, Any]] = None,
         transport: Literal["stdio", "sse", "streamable-http"] = "streamable-http",
+        append_mcp_to_url: bool = True,
         **kwargs,
     ):
         """Initialize MCPToolbox with filtering capabilities.
@@ -45,11 +46,10 @@ class MCPToolbox(MCPTools, metaclass=MCPToolsMeta):
             tool_name (Optional[str], optional): Single tool name to load. Defaults to None.
             headers (Optional[Dict[str, Any]], optional): Headers for toolbox-core client requests. Defaults to None.
             transport (Literal["stdio", "sse", "streamable-http"], optional): MCP transport protocol. Defaults to "streamable-http".
+            append_mcp_to_url (bool, optional): Whether to append "/mcp" to the URL if it doesn't end with it. Defaults to True.
 
         """
-
-        # Ensure the URL ends in "/mcp" as expected
-        if not url.endswith("/mcp"):
+        if append_mcp_to_url and not url.endswith("/mcp"):
             url = url + "/mcp"
 
         super().__init__(url=url, transport=transport, **kwargs)

agno/tools/scrapegraph.py

@@ -187,7 +187,7 @@ class ScrapeGraphTools(Toolkit):
         """
         try:
             log_debug(f"ScrapeGraph searchscraper request with prompt: {user_prompt}")
-            response = self.client.searchscraper(user_prompt=user_prompt, render_heavy_js=self.render_heavy_js)
+            response = self.client.searchscraper(user_prompt=user_prompt)
             return json.dumps(response["result"])
         except Exception as e:
             error_msg = f"Searchscraper failed: {str(e)}"

agno/utils/models/claude.py

@@ -32,6 +32,7 @@ class MCPServerConfiguration:
 
 ROLE_MAP = {
     "system": "system",
+    "developer": "system",
     "user": "user",
     "assistant": "assistant",
     "tool": "user",
@@ -217,7 +218,8 @@ def format_messages(messages: List[Message]) -> Tuple[List[Dict[str, str]], str]
 
     for message in messages:
         content = message.content or ""
-        if message.role == "system":
+        # Both "system" and "developer" roles should be extracted as system messages
+        if message.role in ("system", "developer"):
             if content is not None:
                 system_messages.append(content)  # type: ignore
             continue

agno/utils/streamlit.py

@@ -80,7 +80,7 @@ def session_selector_widget(agent: Agent, model_id: str, agent_creation_callback
     session_options = []
     session_dict = {}
 
-    for session in sessions:
+    for session in sessions:  # type: ignore
         if not hasattr(session, "session_id") or not session.session_id:
             continue
 

agno/vectordb/base.py

@@ -1,11 +1,28 @@
 from abc import ABC, abstractmethod
 from typing import Any, Dict, List, Optional
 
+from agno.knowledge.document import Document
+from agno.utils.string import generate_id
+
 
 class VectorDb(ABC):
     """Base class for Vector Databases"""
 
-    from agno.knowledge.document import Document
+    def __init__(self, *, id: Optional[str] = None, name: Optional[str] = None, description: Optional[str] = None):
+        """Initialize base VectorDb.
+
+        Args:
+            id: Optional custom ID. If not provided, an id will be generated.
+            name: Optional name for the vector database.
+            description: Optional description for the vector database.
+        """
+        if name is None:
+            name = self.__class__.__name__
+
+        self.name = name
+        self.description = description
+        # Last resort fallback to generate id from name if ID not specified
+        self.id = id if id else generate_id(name)
 
     @abstractmethod
     def create(self) -> None:
@@ -106,3 +123,7 @@ class VectorDb(ABC):
     @abstractmethod
     def delete_by_content_id(self, content_id: str) -> bool:
         raise NotImplementedError
+
+    @abstractmethod
+    def get_supported_search_types(self) -> List[str]:
+        raise NotImplementedError

agno/vectordb/cassandra/cassandra.py

@@ -15,6 +15,8 @@ class Cassandra(VectorDb):
         keyspace: str,
         embedder: Optional[Embedder] = None,
         session=None,
+        name: Optional[str] = None,
+        description: Optional[str] = None,
     ) -> None:
         if not table_name:
             raise ValueError("Table name must be provided.")
@@ -30,6 +32,9 @@ class Cassandra(VectorDb):
 
             embedder = OpenAIEmbedder()
             log_info("Embedder not provided, using OpenAIEmbedder as default.")
+        # Initialize base class with name and description
+        super().__init__(name=name, description=description)
+
         self.table_name: str = table_name
         self.embedder: Embedder = embedder
         self.session = session
@@ -483,3 +488,7 @@ class Cassandra(VectorDb):
         except Exception as e:
             log_error(f"Error updating metadata for content_id {content_id}: {e}")
             raise
+
+    def get_supported_search_types(self) -> List[str]:
+        """Get the supported search types for this vector database."""
+        return []  # Cassandra doesn't use SearchType enum

agno/vectordb/chroma/chromadb.py

@@ -25,6 +25,9 @@ class ChromaDb(VectorDb):
     def __init__(
         self,
         collection: str,
+        name: Optional[str] = None,
+        description: Optional[str] = None,
+        id: Optional[str] = None,
         embedder: Optional[Embedder] = None,
         distance: Distance = Distance.cosine,
         path: str = "tmp/chromadb",
@@ -32,9 +35,22 @@ class ChromaDb(VectorDb):
         reranker: Optional[Reranker] = None,
         **kwargs,
     ):
+        # Validate required parameters
+        if not collection:
+            raise ValueError("Collection name must be provided.")
+
+        # Dynamic ID generation based on unique identifiers
+        if id is None:
+            from agno.utils.string import generate_id
+
+            seed = f"{path}#{collection}"
+            id = generate_id(seed)
+
+        # Initialize base class with name, description, and generated ID
+        super().__init__(id=id, name=name, description=description)
+
         # Collection attributes
         self.collection_name: str = collection
-
         # Embedder for embedding the document contents
         if embedder is None:
             from agno.knowledge.embedder.openai import OpenAIEmbedder
@@ -497,11 +513,11 @@ class ChromaDb(VectorDb):
         # Build search results
         search_results: List[Document] = []
 
-        ids_list = result.get("ids", [[]])
-        metadata_list = result.get("metadatas", [[{}]])
-        documents_list = result.get("documents", [[]])
-        embeddings_list = result.get("embeddings")
-        distances_list = result.get("distances", [[]])
+        ids_list = result.get("ids", [[]])  # type: ignore
+        metadata_list = result.get("metadatas", [[{}]])  # type: ignore
+        documents_list = result.get("documents", [[]])  # type: ignore
+        embeddings_list = result.get("embeddings")  # type: ignore
+        distances_list = result.get("distances", [[]])  # type: ignore
 
         if not ids_list or not metadata_list or not documents_list or embeddings_list is None or not distances_list:
             return search_results
@@ -901,3 +917,7 @@ class ChromaDb(VectorDb):
         except Exception as e:
             logger.error(f"Error updating metadata for content_id '{content_id}': {e}")
             raise
+
+    def get_supported_search_types(self) -> List[str]:
+        """Get the supported search types for this vector database."""
+        return []  # ChromaDb doesn't use SearchType enum

agno/vectordb/clickhouse/clickhousedb.py

@@ -23,6 +23,8 @@ class Clickhouse(VectorDb):
         self,
         table_name: str,
         host: str,
+        name: Optional[str] = None,
+        description: Optional[str] = None,
         username: Optional[str] = None,
         password: str = "",
         port: int = 0,
@@ -41,9 +43,11 @@ class Clickhouse(VectorDb):
         self.password = password
         self.port = port
         self.dsn = dsn
+        # Initialize base class with name and description
+        super().__init__(name=name, description=description)
+
         self.compress = compress
         self.database_name = database_name
-
         if not client:
             client = clickhouse_connect.get_client(
                 host=self.host,
@@ -817,3 +821,7 @@ class Clickhouse(VectorDb):
         except Exception as e:
             logger.error(f"Error updating metadata for content_id '{content_id}': {e}")
             raise
+
+    def get_supported_search_types(self) -> List[str]:
+        """Get the supported search types for this vector database."""
+        return []  # Clickhouse doesn't use SearchType enum

agno/vectordb/couchbase/couchbase.py

@@ -66,6 +66,8 @@ class CouchbaseSearch(VectorDb):
         is_global_level_index: bool = False,
         wait_until_index_ready: float = 0,
         batch_limit: int = 500,
+        name: Optional[str] = None,
+        description: Optional[str] = None,
         **kwargs,
     ):
         """
@@ -75,6 +77,8 @@ class CouchbaseSearch(VectorDb):
             bucket_name (str): Name of the Couchbase bucket.
             scope_name (str): Name of the scope within the bucket.
             collection_name (str): Name of the collection within the scope.
+            name (Optional[str]): Name of the vector database.
+            description (Optional[str]): Description of the vector database.
             couchbase_connection_string (str): Couchbase connection string.
             cluster_options (ClusterOptions): Options for configuring the Couchbase cluster connection.
             search_index (Union[str, SearchIndex], optional): Search index configuration, either as index name or SearchIndex definition.
@@ -96,6 +100,9 @@ class CouchbaseSearch(VectorDb):
         self.overwrite = overwrite
         self.is_global_level_index = is_global_level_index
         self.wait_until_index_ready = wait_until_index_ready
+        # Initialize base class with name and description
+        super().__init__(name=name, description=description)
+
         self.kwargs = kwargs
         self.batch_limit = batch_limit
         if isinstance(search_index, str):
@@ -1420,3 +1427,7 @@ class CouchbaseSearch(VectorDb):
         except Exception as e:
             logger.error(f"Error updating metadata for content_id '{content_id}': {e}")
             raise
+
+    def get_supported_search_types(self) -> List[str]:
+        """Get the supported search types for this vector database."""
+        return []  # CouchbaseSearch doesn't use SearchType enum

agno/vectordb/lancedb/lance_db.py

@@ -25,6 +25,8 @@ class LanceDb(VectorDb):
 
     Args:
         uri: The URI of the LanceDB database.
+        name: Name of the vector database.
+        description: Description of the vector database.
         connection: The LanceDB connection to use.
         table: The LanceDB table instance to use.
         async_connection: The LanceDB async connection to use.
@@ -44,6 +46,9 @@ class LanceDb(VectorDb):
     def __init__(
         self,
         uri: lancedb.URI = "/tmp/lancedb",
+        name: Optional[str] = None,
+        description: Optional[str] = None,
+        id: Optional[str] = None,
         connection: Optional[lancedb.LanceDBConnection] = None,
         table: Optional[lancedb.db.LanceTable] = None,
         async_connection: Optional[lancedb.AsyncConnection] = None,
@@ -59,6 +64,17 @@ class LanceDb(VectorDb):
         on_bad_vectors: Optional[str] = None,  # One of "error", "drop", "fill", "null".
         fill_value: Optional[float] = None,  # Only used if on_bad_vectors is "fill"
     ):
+        # Dynamic ID generation based on unique identifiers
+        if id is None:
+            from agno.utils.string import generate_id
+
+            table_identifier = table_name or "default_table"
+            seed = f"{uri}#{table_identifier}"
+            id = generate_id(seed)
+
+        # Initialize base class with name, description, and generated ID
+        super().__init__(id=id, name=name, description=description)
+
         # Embedder for embedding the document contents
         if embedder is None:
             from agno.knowledge.embedder.openai import OpenAIEmbedder
@@ -1048,3 +1064,7 @@ class LanceDb(VectorDb):
         except Exception as e:
             logger.error(f"Error updating metadata for content_id '{content_id}': {e}")
             raise
+
+    def get_supported_search_types(self) -> List[str]:
+        """Get the supported search types for this vector database."""
+        return [SearchType.vector, SearchType.keyword, SearchType.hybrid]

agno/vectordb/langchaindb/langchaindb.py

@@ -11,16 +11,23 @@ class LangChainVectorDb(VectorDb):
         vectorstore: Optional[Any] = None,
         search_kwargs: Optional[dict] = None,
         knowledge_retriever: Optional[Any] = None,
+        name: Optional[str] = None,
+        description: Optional[str] = None,
     ):
         """
         Initialize LangChainVectorDb.
 
         Args:
             vectorstore: The LangChain vectorstore instance
+            name (Optional[str]): Name of the vector database.
+            description (Optional[str]): Description of the vector database.
             search_kwargs: Additional search parameters for the retriever
             knowledge_retriever: An optional LangChain retriever instance
         """
         self.vectorstore = vectorstore
+        # Initialize base class with name and description
+        super().__init__(name=name, description=description)
+
         self.search_kwargs = search_kwargs
         self.knowledge_retriever = knowledge_retriever
 
@@ -141,3 +148,7 @@ class LangChainVectorDb(VectorDb):
             metadata (Dict[str, Any]): The metadata to update
         """
         raise NotImplementedError("update_metadata not supported for LangChain vectorstores")
+
+    def get_supported_search_types(self) -> List[str]:
+        """Get the supported search types for this vector database."""
+        return []  # LangChainVectorDb doesn't use SearchType enum

agno/vectordb/lightrag/lightrag.py

@@ -21,9 +21,14 @@ class LightRag(VectorDb):
         api_key: Optional[str] = None,
         auth_header_name: str = "X-API-KEY",
         auth_header_format: str = "{api_key}",
+        name: Optional[str] = None,
+        description: Optional[str] = None,
     ):
         self.server_url = server_url
         self.api_key = api_key
+        # Initialize base class with name and description
+        super().__init__(name=name, description=description)
+
         self.auth_header_name = auth_header_name
         self.auth_header_format = auth_header_format
 
@@ -372,3 +377,7 @@ class LightRag(VectorDb):
             metadata (Dict[str, Any]): The metadata to update
         """
         raise NotImplementedError("update_metadata not supported for LightRag - use LightRag's native methods")
+
+    def get_supported_search_types(self) -> List[str]:
+        """Get the supported search types for this vector database."""
+        return []  # LightRag doesn't use SearchType enum

agno/vectordb/llamaindex/llamaindexdb.py

@@ -17,8 +17,18 @@ class LlamaIndexVectorDb(VectorDb):
     knowledge_retriever: BaseRetriever
     loader: Optional[Callable] = None
 
-    def __init__(self, knowledge_retriever: BaseRetriever, loader: Optional[Callable] = None, **kwargs):
+    def __init__(
+        self,
+        knowledge_retriever: BaseRetriever,
+        loader: Optional[Callable] = None,
+        name: Optional[str] = None,
+        description: Optional[str] = None,
+        **kwargs,
+    ):
         super().__init__(**kwargs)
+        # Initialize base class with name and description
+        super().__init__(name=name, description=description)
+
         self.knowledge_retriever = knowledge_retriever
         self.loader = loader
 
@@ -144,3 +154,7 @@ class LlamaIndexVectorDb(VectorDb):
             "LlamaIndexVectorDb.delete_by_content_id() not supported - please check the vectorstore manually."
         )
         return False
+
+    def get_supported_search_types(self) -> List[str]:
+        """Get the supported search types for this vector database."""
+        return []  # LlamaIndexVectorDb doesn't use SearchType enum

agno/vectordb/milvus/milvus.py

@@ -28,6 +28,9 @@ class Milvus(VectorDb):
     def __init__(
         self,
         collection: str,
+        name: Optional[str] = None,
+        description: Optional[str] = None,
+        id: Optional[str] = None,
         embedder: Optional[Embedder] = None,
         distance: Distance = Distance.cosine,
         uri: str = "http://localhost:19530",
@@ -42,6 +45,8 @@ class Milvus(VectorDb):
 
         Args:
             collection (str): Name of the Milvus collection.
+            name (Optional[str]): Name of the vector database.
+            description (Optional[str]): Description of the vector database.
             embedder (Embedder): Embedder to use for embedding documents.
             distance (Distance): Distance metric to use for vector similarity.
             uri (Optional[str]): URI of the Milvus server.
@@ -63,6 +68,20 @@ class Milvus(VectorDb):
             reranker (Optional[Reranker]): Reranker to use for hybrid search results
             **kwargs: Additional keyword arguments to pass to the MilvusClient.
         """
+        # Validate required parameters
+        if not collection:
+            raise ValueError("Collection name must be provided.")
+
+        # Dynamic ID generation based on unique identifiers
+        if id is None:
+            from agno.utils.string import generate_id
+
+            seed = f"{uri or 'milvus'}#{collection}"
+            id = generate_id(seed)
+
+        # Initialize base class with name, description, and generated ID
+        super().__init__(id=id, name=name, description=description)
+
         self.collection: str = collection
 
         if embedder is None:
@@ -1141,3 +1160,7 @@ class Milvus(VectorDb):
         except Exception as e:
             log_error(f"Error updating metadata for content_id '{content_id}': {e}")
             raise
+
+    def get_supported_search_types(self) -> List[str]:
+        """Get the supported search types for this vector database."""
+        return [SearchType.vector, SearchType.hybrid]

agno/vectordb/mongodb/mongodb.py

@@ -33,6 +33,9 @@ class MongoDb(VectorDb):
     def __init__(
         self,
         collection_name: str,
+        name: Optional[str] = None,
+        description: Optional[str] = None,
+        id: Optional[str] = None,
         db_url: Optional[str] = "mongodb://localhost:27017/",
         database: str = "agno",
         embedder: Optional[Embedder] = None,
@@ -56,6 +59,8 @@ class MongoDb(VectorDb):
 
         Args:
             collection_name (str): Name of the MongoDB collection.
+            name (Optional[str]): Name of the vector database.
+            description (Optional[str]): Description of the vector database.
             db_url (Optional[str]): MongoDB connection string.
             database (str): Database name.
             embedder (Embedder): Embedder instance for generating embeddings.
@@ -74,11 +79,24 @@ class MongoDb(VectorDb):
             hybrid_rank_constant (int): Default rank constant (k) for Reciprocal Rank Fusion in hybrid search. This constant is added to the rank before taking the reciprocal, helping to smooth scores. A common value is 60.
             **kwargs: Additional arguments for MongoClient.
         """
+        # Validate required parameters
         if not collection_name:
             raise ValueError("Collection name must not be empty.")
         if not database:
             raise ValueError("Database name must not be empty.")
+
+        # Dynamic ID generation based on unique identifiers
+        if id is None:
+            from agno.utils.string import generate_id
+
+            connection_identifier = db_url or "mongodb://localhost:27017/"
+            seed = f"{connection_identifier}#{database}#{collection_name}"
+            id = generate_id(seed)
+
         self.collection_name = collection_name
+        # Initialize base class with name, description, and generated ID
+        super().__init__(id=id, name=name, description=description)
+
         self.database = database
         self.search_index_name = search_index_name
         self.cosmos_compatibility = cosmos_compatibility
@@ -1382,3 +1400,7 @@ class MongoDb(VectorDb):
         except Exception as e:
             logger.error(f"Error updating metadata for content_id '{content_id}': {e}")
             raise
+
+    def get_supported_search_types(self) -> List[str]:
+        """Get the supported search types for this vector database."""
+        return [SearchType.vector, SearchType.hybrid]

agno/vectordb/pgvector/pgvector.py

@@ -3,6 +3,8 @@ from hashlib import md5
 from math import sqrt
 from typing import Any, Dict, List, Optional, Union, cast
 
+from agno.utils.string import generate_id
+
 try:
     from sqlalchemy import update
     from sqlalchemy.dialects import postgresql
@@ -43,6 +45,9 @@ class PgVector(VectorDb):
         self,
         table_name: str,
         schema: str = "ai",
+        name: Optional[str] = None,
+        description: Optional[str] = None,
+        id: Optional[str] = None,
         db_url: Optional[str] = None,
         db_engine: Optional[Engine] = None,
         embedder: Optional[Embedder] = None,
@@ -62,6 +67,8 @@ class PgVector(VectorDb):
         Args:
             table_name (str): Name of the table to store vector data.
             schema (str): Database schema name.
+            name (Optional[str]): Name of the vector database.
+            description (Optional[str]): Description of the vector database.
             db_url (Optional[str]): Database connection URL.
             db_engine (Optional[Engine]): SQLAlchemy database engine.
             embedder (Optional[Embedder]): Embedder instance for creating embeddings.
@@ -80,6 +87,15 @@ class PgVector(VectorDb):
         if db_engine is None and db_url is None:
             raise ValueError("Either 'db_url' or 'db_engine' must be provided.")
 
+        if id is None:
+            base_seed = db_url or str(db_engine.url)  # type: ignore
+            schema_suffix = table_name if table_name is not None else "ai"
+            seed = f"{base_seed}#{schema_suffix}"
+            id = generate_id(seed)
+
+        # Initialize base class with name and description
+        super().__init__(id=id, name=name, description=description)
+
         if db_engine is None:
             if db_url is None:
                 raise ValueError("Must provide 'db_url' if 'db_engine' is None.")
@@ -1383,3 +1399,6 @@ class PgVector(VectorDb):
         copied_obj.table = copied_obj.get_table()
 
         return copied_obj
+
+    def get_supported_search_types(self) -> List[str]:
+        return [SearchType.vector, SearchType.keyword, SearchType.hybrid]