graphiti-core 0.5.3__tar.gz → 0.6.0__tar.gz

{graphiti_core-0.5.3 → graphiti_core-0.6.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: graphiti-core
-Version: 0.5.3
+Version: 0.6.0
 Summary: A temporal graph building library
 License: Apache-2.0
 Author: Paul Paliychuk
@@ -22,7 +22,7 @@ Description-Content-Type: text/markdown
 
 <div align="center">
 
-<img width="350" alt="Graphiti-ts-small" src="https://github.com/user-attachments/assets/bbd02947-e435-4a05-b25a-bbbac36d52c8">
+# Graphiti
 
 ## Temporal Knowledge Graphs for Agentic Applications
 
@@ -45,7 +45,7 @@ a fusion of time, full-text, semantic, and graph algorithm approaches.
 <br />
 
 <p align="center">
-    <img src="/images/graphiti-graph-intro.gif" alt="Graphiti temporal walkthrough" width="700px">   
+    <img src="images/graphiti-graph-intro.gif" alt="Graphiti temporal walkthrough" width="700px">   
 </p>
 
 <br />
@@ -65,6 +65,20 @@ With Graphiti, you can build LLM applications such as:
 Graphiti supports a wide range of applications in sales, customer service, health, finance, and more, enabling long-term
 recall and state-based reasoning for both assistants and agents.
 
+## Graphiti and Zep Memory
+
+Graphiti powers the core of [Zep's memory layer](https://www.getzep.com) for LLM-powered Assistants and Agents.
+
+Using Graphiti, we've demonstrated Zep is the [State of the Art in Agent Memory](https://blog.getzep.com/state-of-the-art-agent-memory/).
+
+Read our paper: [Zep: A Temporal Knowledge Graph Architecture for Agent Memory](https://arxiv.org/abs/2501.13956).
+
+We're excited to open-source Graphiti, believing its potential reaches far beyond memory applications.
+
+<p align="center">
+    <a href="https://arxiv.org/abs/2501.13956"><img src="images/arxiv-screenshot.png" alt="Zep: A Temporal Knowledge Graph Architecture for Agent Memory" width="700px"></a>
+</p>
+
 ## Why Graphiti?
 
 We were intrigued by Microsoft’s GraphRAG, which expanded on RAG text chunking by using a graph to better model a
@@ -89,12 +103,6 @@ scale:
     <img src="/images/graphiti-intro-slides-stock-2.gif" alt="Graphiti structured + unstructured demo" width="700px">   
 </p>
 
-## Graphiti and Zep Memory
-
-Graphiti powers the core of [Zep's memory layer](https://www.getzep.com) for LLM-powered Assistants and Agents.
-
-We're excited to open-source Graphiti, believing its potential reaches far beyond memory applications.
-
 ## Installation
 
 Requirements:
@@ -125,7 +133,7 @@ poetry add graphiti-core
 
 > [!IMPORTANT]
 > Graphiti uses OpenAI for LLM inference and embedding. Ensure that an `OPENAI_API_KEY` is set in your environment.
-> Support for Anthropic and Groq LLM inferences is available, too.
+> Support for Anthropic and Groq LLM inferences is available, too. Other LLM providers may be supported via OpenAI compatible APIs.
 
 ```python
 from graphiti_core import Graphiti
@@ -216,13 +224,9 @@ as such this feature is off by default.
 
 Graphiti is under active development. We aim to maintain API stability while working on:
 
-- [x] Implementing node and edge CRUD operations
-- [ ] Improving performance and scalability
-- [ ] Achieving good performance with different LLM and embedding models
-- [x] Creating a dedicated embedder interface
 - [ ] Supporting custom graph schemas:
-    - Allow developers to provide their own defined node and edge classes when ingesting episodes
-    - Enable more flexible knowledge representation tailored to specific use cases
+  - Allow developers to provide their own defined node and edge classes when ingesting episodes
+  - Enable more flexible knowledge representation tailored to specific use cases
 - [x] Enhancing retrieval capabilities with more robust and configurable options
 - [ ] Expanding test coverage to ensure reliability and catch edge cases
 

{graphiti_core-0.5.3 → graphiti_core-0.6.0}/README.md

@@ -1,6 +1,6 @@
 <div align="center">
 
-<img width="350" alt="Graphiti-ts-small" src="https://github.com/user-attachments/assets/bbd02947-e435-4a05-b25a-bbbac36d52c8">
+# Graphiti
 
 ## Temporal Knowledge Graphs for Agentic Applications
 
@@ -23,7 +23,7 @@ a fusion of time, full-text, semantic, and graph algorithm approaches.
 <br />
 
 <p align="center">
-    <img src="/images/graphiti-graph-intro.gif" alt="Graphiti temporal walkthrough" width="700px">   
+    <img src="images/graphiti-graph-intro.gif" alt="Graphiti temporal walkthrough" width="700px">   
 </p>
 
 <br />
@@ -43,6 +43,20 @@ With Graphiti, you can build LLM applications such as:
 Graphiti supports a wide range of applications in sales, customer service, health, finance, and more, enabling long-term
 recall and state-based reasoning for both assistants and agents.
 
+## Graphiti and Zep Memory
+
+Graphiti powers the core of [Zep's memory layer](https://www.getzep.com) for LLM-powered Assistants and Agents.
+
+Using Graphiti, we've demonstrated Zep is the [State of the Art in Agent Memory](https://blog.getzep.com/state-of-the-art-agent-memory/).
+
+Read our paper: [Zep: A Temporal Knowledge Graph Architecture for Agent Memory](https://arxiv.org/abs/2501.13956).
+
+We're excited to open-source Graphiti, believing its potential reaches far beyond memory applications.
+
+<p align="center">
+    <a href="https://arxiv.org/abs/2501.13956"><img src="images/arxiv-screenshot.png" alt="Zep: A Temporal Knowledge Graph Architecture for Agent Memory" width="700px"></a>
+</p>
+
 ## Why Graphiti?
 
 We were intrigued by Microsoft’s GraphRAG, which expanded on RAG text chunking by using a graph to better model a
@@ -67,12 +81,6 @@ scale:
     <img src="/images/graphiti-intro-slides-stock-2.gif" alt="Graphiti structured + unstructured demo" width="700px">   
 </p>
 
-## Graphiti and Zep Memory
-
-Graphiti powers the core of [Zep's memory layer](https://www.getzep.com) for LLM-powered Assistants and Agents.
-
-We're excited to open-source Graphiti, believing its potential reaches far beyond memory applications.
-
 ## Installation
 
 Requirements:
@@ -103,7 +111,7 @@ poetry add graphiti-core
 
 > [!IMPORTANT]
 > Graphiti uses OpenAI for LLM inference and embedding. Ensure that an `OPENAI_API_KEY` is set in your environment.
-> Support for Anthropic and Groq LLM inferences is available, too.
+> Support for Anthropic and Groq LLM inferences is available, too. Other LLM providers may be supported via OpenAI compatible APIs.
 
 ```python
 from graphiti_core import Graphiti
@@ -194,13 +202,9 @@ as such this feature is off by default.
 
 Graphiti is under active development. We aim to maintain API stability while working on:
 
-- [x] Implementing node and edge CRUD operations
-- [ ] Improving performance and scalability
-- [ ] Achieving good performance with different LLM and embedding models
-- [x] Creating a dedicated embedder interface
 - [ ] Supporting custom graph schemas:
-    - Allow developers to provide their own defined node and edge classes when ingesting episodes
-    - Enable more flexible knowledge representation tailored to specific use cases
+  - Allow developers to provide their own defined node and edge classes when ingesting episodes
+  - Enable more flexible knowledge representation tailored to specific use cases
 - [x] Enhancing retrieval capabilities with more robust and configurable options
 - [ ] Expanding test coverage to ensure reliability and catch edge cases
 

{graphiti_core-0.5.3 → graphiti_core-0.6.0}/graphiti_core/graphiti.py

@@ -21,6 +21,7 @@ from time import time
 from dotenv import load_dotenv
 from neo4j import AsyncGraphDatabase
 from pydantic import BaseModel
+from typing_extensions import LiteralString
 
 from graphiti_core.cross_encoder.client import CrossEncoderClient
 from graphiti_core.cross_encoder.openai_reranker_client import OpenAIRerankerClient
@@ -747,3 +748,34 @@ class Graphiti:
         await add_nodes_and_edges_bulk(
             self.driver, [], [], resolved_nodes, [resolved_edge] + invalidated_edges
         )
+
+    async def remove_episode(self, episode_uuid: str):
+        # Find the episode to be deleted
+        episode = await EpisodicNode.get_by_uuid(self.driver, episode_uuid)
+
+        # Find edges mentioned by the episode
+        edges = await EntityEdge.get_by_uuids(self.driver, episode.entity_edges)
+
+        # We should only delete edges created by the episode
+        edges_to_delete: list[EntityEdge] = []
+        for edge in edges:
+            if edge.episodes[0] == episode.uuid:
+                edges_to_delete.append(edge)
+
+        # Find nodes mentioned by the episode
+        nodes = await get_mentioned_nodes(self.driver, [episode])
+        # We should delete all nodes that are only mentioned in the deleted episode
+        nodes_to_delete: list[EntityNode] = []
+        for node in nodes:
+            query: LiteralString = 'MATCH (e:Episodic)-[:MENTIONS]->(n:Entity {uuid: $uuid}) RETURN count(*) AS episode_count'
+            records, _, _ = await self.driver.execute_query(
+                query, uuid=node.uuid, database_=DEFAULT_DATABASE, routing_='r'
+            )
+
+            for record in records:
+                if record['episode_count'] == 1:
+                    nodes_to_delete.append(node)
+
+        await semaphore_gather(*[node.delete(self.driver) for node in nodes_to_delete])
+        await semaphore_gather(*[edge.delete(self.driver) for edge in edges_to_delete])
+        await episode.delete(self.driver)

{graphiti_core-0.5.3 → graphiti_core-0.6.0}/graphiti_core/nodes.py

@@ -170,7 +170,8 @@ class EpisodicNode(Node):
             e.name AS name,
             e.group_id AS group_id,
             e.source_description AS source_description,
-            e.source AS source
+            e.source AS source,
+            e.entity_edges AS entity_edges
         """,
             uuid=uuid,
             database_=DEFAULT_DATABASE,
@@ -197,7 +198,8 @@ class EpisodicNode(Node):
             e.name AS name,
             e.group_id AS group_id,
             e.source_description AS source_description,
-            e.source AS source
+            e.source AS source,
+            e.entity_edges AS entity_edges
         """,
             uuids=uuids,
             database_=DEFAULT_DATABASE,
@@ -233,7 +235,8 @@ class EpisodicNode(Node):
             e.name AS name,
             e.group_id AS group_id,
             e.source_description AS source_description,
-            e.source AS source
+            e.source AS source,
+            e.entity_edges AS entity_edges
         ORDER BY e.uuid DESC
         """
             + limit_query,
@@ -490,6 +493,7 @@ def get_episodic_node_from_record(record: Any) -> EpisodicNode:
         source=EpisodeType.from_str(record['source']),
         name=record['name'],
         source_description=record['source_description'],
+        entity_edges=record['entity_edges'],
     )
 
 

{graphiti_core-0.5.3 → graphiti_core-0.6.0}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "graphiti-core"
-version = "0.5.3"
+version = "0.6.0"
 description = "A temporal graph building library"
 authors = [
     "Paul Paliychuk <paul@getzep.com>",