zep: a suite of integrations and tools for implementing long-term agent memory via Zep Cloud

What it solves

It provides a set of tools, example code, and framework integrations to help developers implement long-term memory for AI agents using Zep Cloud, a managed agent memory platform.

How it works

This repository acts as a hub for implementing Zep Cloud's capabilities. It includes SDKs for Python, TypeScript/JavaScript, and Go, as well as a bulk data ingestion pipeline (zep-ingest) for importing data from sources like Slack, email, and documents. It also provides pre-built integration packages for popular agent frameworks such as CrewAI, LangGraph, AutoGen, and the Vercel AI SDK.

Who it’s for

Developers building AI agents who need a scalable way to manage agent memory and integrate it with various orchestration frameworks.

Highlights

• Broad Framework Support: Integrations for a wide range of Python, TypeScript, and Go frameworks including Pydantic AI and Microsoft AutoGen.
• Bulk Ingestion: A dedicated pipeline for importing data from multiple formats including JSON, CSV, and fact triples.
• Evaluation Tools: Includes a memory benchmark suite (LoCoMo, LongMemEval) and an evaluation harness for retrieval and ingestion.
• Multi-language SDKs: Official support for Python, TypeScript/JavaScript, and Go.