Zep is an end-to-end context engineering platform that delivers the right information at the right time with sub-200ms latency. It solves the agent context problem by assembling comprehensive, relationship-aware context from multiple data sources—chat history, business data, documents, and app events—enabling AI agents to perform accurately in production.

1. Add context: Feed chat messages, business data, and events to Zep as they occur
2. Graph RAG: Zep automatically extracts relationships and maintains a temporal knowledge graph that understands how context evolves over time
3. Retrieve & assemble: Get pre-formatted, relationship-aware context blocks optimized for your LLM

Zep's relationship-aware retrieval system delivers context about facts, relationships, and how they've changed—providing agents with current, accurate, and relevant information for better decision-making.

Visit www.getzep.com to sign up for Zep Cloud, our managed service delivering intelligent agent context with <200ms latency, enterprise-grade scalability, and SOC2 Type 2 / HIPAA compliance. Add context assembly to your agents in three lines of code.

Zep offers comprehensive SDKs for multiple languages:

• Python: pip install zep-cloud
• TypeScript/JavaScript: npm install @getzep/zep-cloud
• Go: go get github.com/getzep/zep-go/v2

• Documentation: help.getzep.com
• Discord Community: Join our Discord
• Support: Visit our help website for comprehensive guides and tutorials

Note: This repository is currently a work in progress.

This repository contains examples, integrations, and tools for building intelligent agent context with Zep. Explore the example applications to see how Zep delivers context assembly with LangChain, LlamaIndex, AutoGen, and other frameworks.

The repository includes:

• Example applications demonstrating agent context assembly with Zep
• Integration packages for popular agent frameworks
• Code samples for different use cases
• Development tools and utilities

This project uses UV for Python package management with workspace features.

• Python 3.13+
• UV package manager

1. Install UV (if not already installed):

   # On macOS and Linux
   curl -LsSf https://astral.sh/uv/install.sh | sh
   
   # On Windows
   powershell -c "irm https://astral.sh/uv/install.ps1 | iex"

2. Sync the workspace:

   uv sync

3. Activate the virtual environment:

   # On Unix/macOS
   source .venv/bin/activate
   
   # On Windows
   .venv\Scripts\activate

This project is configured as a UV workspace with the following packages:

• Main package (zep): Core Zep functionality
• Autogen integration (integration/autogen/): Integration package for Autogen framework

• Install dependencies: uv sync
• Add dependencies to main package: uv add <package>
• Add dependencies to autogen integration: uv add --project integration/autogen <package>
• Run tests for autogen integration: uv run --project integration/autogen pytest
• Build packages: uv build

The autogen integration package is located in integration/autogen/ with the following structure:

integration/autogen/
├── src/zep_autogen/     # Package source code
├── tests/               # Test files
└── pyproject.toml       # Package configuration

We welcome contributions to help improve Zep and its ecosystem. Please see the CONTRIBUTING.md file for guidelines on how to contribute, including:

• Code contributions
• Documentation improvements
• Bug reports and feature requests
• Community examples and integrations

Zep is powered by Graphiti, an open-source temporal knowledge graph framework that enables relationship-aware context retrieval.

Graphiti autonomously builds and maintains knowledge graphs while reasoning about state changes over time. Each fact includes valid_at and invalid_at dates, allowing agents to understand how relationships, preferences, and context have evolved—essential for accurate decision-making in dynamic environments.

Visit the Graphiti repository to learn more about the temporal knowledge graph framework that powers Zep's context assembly capabilities.

Note: Zep Community Edition is no longer supported and has been deprecated. The Community Edition code has been moved to the legacy/ folder in this repository.

For current Zep development, we recommend using Zep Cloud or exploring the example projects in this repository.

Read more about this change in our announcement: Announcing a New Direction for Zep's Open Source Strategy