cognee 1.0.1.dev0__tar.gz → 1.0.1.dev2__tar.gz

{cognee-1.0.1.dev0 → cognee-1.0.1.dev2}/.env.template

@@ -281,6 +281,10 @@ TOKENIZERS_PARALLELISM="false"
 # Number of rotated log files to keep (default: 5 → 300 MB total cap)
 #COGNEE_LOG_BACKUP_COUNT=5
 
+# -- Search History ------------------------------------------------------------
+# Set to false to disable search query/result logging (recommended for daemons)
+#COGNEE_LOG_SEARCH_HISTORY="true"
+
 # LITELLM Logging Level. Set to quiet down logging
 LITELLM_LOG="ERROR"
 #TELEMETRY_DISABLED=1

{cognee-1.0.1.dev0 → cognee-1.0.1.dev2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: cognee
-Version: 1.0.1.dev0
+Version: 1.0.1.dev2
 Summary: Cognee - is a library for enriching LLM context with a semantic layer for better understanding and reasoning.
 Project-URL: Homepage, https://www.cognee.ai
 Project-URL: Repository, https://github.com/topoteretes/cognee

cognee-1.0.1.dev2/cognee/alembic/versions/b2c3d4e5f6a7_add_search_history_indexes.py

@@ -0,0 +1,44 @@
+"""add created_at index to queries and results tables
+
+Revision ID: b2c3d4e5f6a7
+Revises:
+Create Date: 2026-04-04
+"""
+
+from alembic import op
+from sqlalchemy.engine.reflection import Inspector
+
+revision = "b2c3d4e5f6a7"
+down_revision = "6ee94db7381f"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    conn = op.get_bind()
+    inspector = Inspector.from_engine(conn)
+
+    # Get existing indexes for 'queries'
+    queries_indexes = [idx["name"] for idx in inspector.get_indexes("queries")]
+    if "ix_queries_created_at" not in queries_indexes:
+        op.create_index("ix_queries_created_at", "queries", ["created_at"])
+
+    # Get existing indexes for 'results'
+    results_indexes = [idx["name"] for idx in inspector.get_indexes("results")]
+    if "ix_results_created_at" not in results_indexes:
+        op.create_index("ix_results_created_at", "results", ["created_at"])
+
+
+def downgrade() -> None:
+    conn = op.get_bind()
+    inspector = Inspector.from_engine(conn)
+
+    # Get existing indexes for 'results'
+    results_indexes = [idx["name"] for idx in inspector.get_indexes("results")]
+    if "ix_results_created_at" in results_indexes:
+        op.drop_index("ix_results_created_at", "results")
+
+    # Get existing indexes for 'queries'
+    queries_indexes = [idx["name"] for idx in inspector.get_indexes("queries")]
+    if "ix_queries_created_at" in queries_indexes:
+        op.drop_index("ix_queries_created_at", "queries")

{cognee-1.0.1.dev0 → cognee-1.0.1.dev2}/cognee/api/v1/add/add.py

@@ -181,6 +181,17 @@ async def add(
         - TAVILY_API_KEY: YOUR_TAVILY_API_KEY
 
     """
+    # Route to remote instance if connected via serve()
+    from cognee.api.v1.serve.state import get_remote_client
+
+    client = get_remote_client()
+    if client is not None:
+        result = await client.add(data, dataset_name)
+        # Wrap in a simple namespace so callers expecting .model_dump() still work
+        from types import SimpleNamespace
+
+        return SimpleNamespace(**result)
+
     if preferred_loaders is not None:
         transformed = {}
         for item in preferred_loaders:

{cognee-1.0.1.dev0 → cognee-1.0.1.dev2}/cognee/api/v1/cognify/cognify.py

@@ -197,6 +197,13 @@ async def cognify(
         - LLM_RATE_LIMIT_ENABLED: Enable rate limiting (default: False)
         - LLM_RATE_LIMIT_REQUESTS: Max requests per interval (default: 60)
     """
+    # Route to remote instance if connected via serve()
+    from cognee.api.v1.serve.state import get_remote_client
+
+    client = get_remote_client()
+    if client is not None:
+        return await client.cognify(datasets)
+
     with new_span("cognee.api.cognify") as span:
         span.set_attribute(COGNEE_PIPELINE_NAME, "cognify")
         if datasets is not None:

{cognee-1.0.1.dev0 → cognee-1.0.1.dev2}/cognee/api/v1/ontologies/ontologies.py

@@ -7,6 +7,8 @@ from typing import Optional, List
 from dataclasses import dataclass
 from fastapi import UploadFile
 
+from cognee.base_config import get_base_config
+
 
 @dataclass
 class OntologyMetadata:
@@ -23,10 +25,11 @@ class OntologyService:
 
     @property
     def base_dir(self) -> Path:
-        return Path(tempfile.gettempdir()) / "ontologies"
+        base_config = get_base_config()
+        return base_config.data_root_directory
 
     def _get_user_dir(self, user_id: str) -> Path:
-        user_dir = self.base_dir / str(user_id)
+        user_dir = Path(self.base_dir) / str(user_id)
         user_dir.mkdir(parents=True, exist_ok=True)
         return user_dir
 

{cognee-1.0.1.dev0 → cognee-1.0.1.dev2}/cognee/api/v1/search/search.py

@@ -202,6 +202,13 @@ async def search(
         - GRAPH_DATABASE_PROVIDER: Must match what was used during cognify
 
     """
+    # Route to remote instance if connected via serve()
+    from cognee.api.v1.serve.state import get_remote_client
+
+    client = get_remote_client()
+    if client is not None:
+        return await client.search(query_text, search_type=query_type, datasets=datasets)
+
     with new_span("cognee.api.search") as span:
         span.set_attribute(COGNEE_SEARCH_QUERY, query_text[:500])
         span.set_attribute(COGNEE_SEARCH_TYPE, str(query_type.value))

{cognee-1.0.1.dev0 → cognee-1.0.1.dev2}/cognee/api/v1/serve/cloud_client.py

@@ -136,6 +136,83 @@ class CloudClient:
                 raise RuntimeError(f"Remote improve failed ({resp.status}): {body}")
             return await resp.json()
 
+    # ----- V1 Operations (add / cognify / search) -----
+
+    async def add(self, data: Any, dataset_name: str = "main_dataset", **kwargs) -> dict:
+        """POST /api/v1/add — ingest data into a dataset."""
+        session = await self._get_session()
+
+        form = aiohttp.FormData()
+        form.add_field("datasetName", dataset_name)
+
+        if isinstance(data, str):
+            form.add_field(
+                "data",
+                io.BytesIO(data.encode("utf-8")),
+                filename="data.txt",
+                content_type="text/plain",
+            )
+        elif isinstance(data, list):
+            for item in data:
+                if isinstance(item, str):
+                    form.add_field(
+                        "data",
+                        io.BytesIO(item.encode("utf-8")),
+                        filename="data.txt",
+                        content_type="text/plain",
+                    )
+                elif hasattr(item, "read"):
+                    name = getattr(item, "name", "upload")
+                    form.add_field("data", item, filename=name)
+        elif hasattr(data, "read"):
+            name = getattr(data, "name", "upload")
+            form.add_field("data", data, filename=name)
+
+        async with session.post(f"{self.service_url}/api/v1/add", data=form) as resp:
+            if resp.status >= 400:
+                body = await resp.text()
+                raise RuntimeError(f"Remote add failed ({resp.status}): {body}")
+            return await resp.json()
+
+    async def cognify(self, datasets: Any = None, **kwargs) -> dict:
+        """POST /api/v1/cognify — build the knowledge graph."""
+        session = await self._get_session()
+
+        payload: dict = {}
+        if datasets:
+            payload["datasets"] = (
+                [str(d) for d in datasets] if isinstance(datasets, list) else [str(datasets)]
+            )
+
+        async with session.post(
+            f"{self.service_url}/api/v1/cognify",
+            json=payload,
+        ) as resp:
+            if resp.status >= 400:
+                body = await resp.text()
+                raise RuntimeError(f"Remote cognify failed ({resp.status}): {body}")
+            return await resp.json()
+
+    async def search(self, query: str, **kwargs) -> list:
+        """POST /api/v1/search — query the knowledge graph."""
+        session = await self._get_session()
+
+        payload: dict = {"query": query}
+        if kwargs.get("search_type"):
+            st = kwargs["search_type"]
+            payload["searchType"] = st if isinstance(st, str) else st.value
+        if kwargs.get("datasets"):
+            payload["datasets"] = kwargs["datasets"]
+
+        async with session.post(
+            f"{self.service_url}/api/v1/search",
+            json=payload,
+        ) as resp:
+            if resp.status >= 400:
+                body = await resp.text()
+                raise RuntimeError(f"Remote search failed ({resp.status}): {body}")
+            return await resp.json()
+
     async def forget(self, **kwargs) -> dict:
         """POST /api/v1/forget — delete data from the knowledge graph."""
         session = await self._get_session()

{cognee-1.0.1.dev0 → cognee-1.0.1.dev2}/cognee/infrastructure/databases/relational/sqlalchemy/SqlAlchemyAdapter.py

@@ -82,20 +82,15 @@ class SQLAlchemyAdapter:
             # Note: For caching purposes, pool_args is stored as a sorted tuple of key-value pairs in the config
             pool_args = pool_args or {}
 
-            if pool_args.get("pool_size") is None:
-                pool_args["pool_size"] = 20
-
-            if pool_args.get("max_overflow") is None:
-                pool_args["max_overflow"] = 20
-
-            if pool_args.get("pool_pre_ping") is None:
-                pool_args["pool_pre_ping"] = True
-
-            if pool_args.get("pool_recycle") is None:
-                pool_args["pool_recycle"] = 280
-
-            if pool_args.get("pool_timeout") is None:
-                pool_args["pool_timeout"] = 280
+            if pool_args.get("poolclass", "").lower() == "nullpool":
+                pool_args["poolclass"] = NullPool
+            else:
+                # Standard QueuePool settings
+                pool_args.setdefault("pool_size", 20)
+                pool_args.setdefault("max_overflow", 20)
+                pool_args.setdefault("pool_pre_ping", True)
+                pool_args.setdefault("pool_recycle", 280)
+                pool_args.setdefault("pool_timeout", 280)
 
             engine_kwargs = {**pool_args}
             if final_connect_args:

{cognee-1.0.1.dev0 → cognee-1.0.1.dev2}/cognee/modules/search/methods/search.py

@@ -1,7 +1,6 @@
 import json
 import asyncio
 from uuid import UUID
-from fastapi.encoders import jsonable_encoder
 from typing import Any, List, Optional, Tuple, Type, Union
 
 from cognee.infrastructure.databases.graph import get_graph_engine
@@ -122,9 +121,19 @@ async def search(
         },
     )
 
+    # Log only the completion text (what the user sees), not the full
+    # serialized graph payload. The raw result_objects can be 50-100 KB
+    # each and cause unbounded DB growth in long-running deployments.
+    completions = []
+    for item in search_results:
+        payload = item[0] if isinstance(item, tuple) else item
+        if hasattr(payload, "completion") and payload.completion:
+            completions.append(payload.completion)
+        elif hasattr(payload, "context") and payload.context:
+            completions.append(payload.context)
     await log_result(
         query.id,
-        json.dumps(jsonable_encoder(search_results)),
+        json.dumps(completions) if completions else "[]",
         user.id,
     )
 

{cognee-1.0.1.dev0 → cognee-1.0.1.dev2}/cognee/modules/search/models/Query.py

@@ -11,7 +11,9 @@ class Query(Base):
 
     text = Column(String)
     query_type = Column(String)
-    user_id = Column(UUID)
+    user_id = Column(UUID, index=True)
 
-    created_at = Column(DateTime(timezone=True), default=lambda: datetime.now(timezone.utc))
+    created_at = Column(
+        DateTime(timezone=True), default=lambda: datetime.now(timezone.utc), index=True
+    )
     updated_at = Column(DateTime(timezone=True), onupdate=lambda: datetime.now(timezone.utc))

{cognee-1.0.1.dev0 → cognee-1.0.1.dev2}/cognee/modules/search/models/Result.py

@@ -13,5 +13,7 @@ class Result(Base):
     query_id = Column(UUID)
     user_id = Column(UUID, index=True)
 
-    created_at = Column(DateTime(timezone=True), default=lambda: datetime.now(timezone.utc))
+    created_at = Column(
+        DateTime(timezone=True), default=lambda: datetime.now(timezone.utc), index=True
+    )
     updated_at = Column(DateTime(timezone=True), onupdate=lambda: datetime.now(timezone.utc))

{cognee-1.0.1.dev0 → cognee-1.0.1.dev2}/cognee/modules/search/operations/log_query.py

@@ -1,20 +1,24 @@
+import os
 from uuid import UUID
 from cognee.infrastructure.databases.relational import get_relational_engine
 from ..models.Query import Query
 
+_LOG_ENABLED = os.getenv("COGNEE_LOG_SEARCH_HISTORY", "true").lower() in ("true", "1", "yes")
+
 
 async def log_query(query_text: str, query_type: str, user_id: UUID) -> Query:
+    query = Query(
+        text=query_text,
+        query_type=query_type,
+        user_id=user_id,
+    )
+
+    if not _LOG_ENABLED:
+        return query
+
     db_engine = get_relational_engine()
 
     async with db_engine.get_async_session() as session:
-        query = Query(
-            text=query_text,
-            query_type=query_type,
-            user_id=user_id,
-        )
-
         session.add(query)
-
         await session.commit()
-
         return query

{cognee-1.0.1.dev0 → cognee-1.0.1.dev2}/cognee/modules/search/operations/log_result.py

@@ -1,9 +1,15 @@
+import os
 from uuid import UUID
 from cognee.infrastructure.databases.relational import get_relational_engine
 from ..models.Result import Result
 
+_LOG_ENABLED = os.getenv("COGNEE_LOG_SEARCH_HISTORY", "true").lower() in ("true", "1", "yes")
+
 
 async def log_result(query_id: UUID, result: str, user_id: UUID):
+    if not _LOG_ENABLED:
+        return
+
     db_engine = get_relational_engine()
 
     async with db_engine.get_async_session() as session:

{cognee-1.0.1.dev0 → cognee-1.0.1.dev2}/cognee/tests/test_cognee_server_start.py

@@ -28,7 +28,7 @@ class TestCogneeServerStart(unittest.TestCase):
             preexec_fn=os.setsid,
         )
         # Give the server some time to start
-        time.sleep(35)
+        time.sleep(120)
 
         # Check if server started with errors
         if cls.server_process.poll() is not None:

{cognee-1.0.1.dev0 → cognee-1.0.1.dev2}/pyproject.toml

@@ -1,7 +1,7 @@
 [project]
 name = "cognee"
 
-version = "1.0.1.dev0"
+version = "1.0.1.dev2"
 description = "Cognee - is a library for enriching LLM context with a semantic layer for better understanding and reasoning."
 authors = [
     { name = "Vasilije Markovic" },

{cognee-1.0.1.dev0 → cognee-1.0.1.dev2}/uv.lock

@@ -1053,7 +1053,7 @@ wheels = [
 
 [[package]]
 name = "cognee"
-version = "1.0.1.dev0"
+version = "1.0.1.dev2"
 source = { editable = "." }
 dependencies = [
     { name = "aiofiles" },