Claude-Mem

Claude-Mem is an open-source memory engine with 102,000+ stars on GitHub that gives AI coding agents persistent context across sessions, so they stop forgetting your project between conversations.

Overview

Every new agent session normally starts from zero: the agent rediscovers your codebase, your conventions and your past decisions the hard way. Claude-Mem fixes that by capturing everything your agent does during a session — files touched, commands run, decisions made — compressing it with AI into a searchable memory, and injecting the relevant context back into future sessions automatically. It works as a plugin across the agent ecosystem: Claude Code, OpenClaw, Codex, Gemini, Hermes, Copilot, OpenCode and more. Under the hood it combines SQLite storage with ChromaDB vector search and embedding-based retrieval, so recall is semantic rather than a raw transcript dump. The project is Apache-2.0 licensed and extremely active, with repository updates in October 2026, and its six-figure star count shows how universal the agent-memory problem has become.

Key Features

• Automatic capture – records agent activity during sessions without manual note-taking.
• AI compression – sessions are distilled into dense, useful memories instead of raw logs.
• Context injection – relevant memories surface in new sessions at the right moment.
• Semantic search – ChromaDB embeddings find memories by meaning, not keywords.
• Agent-agnostic – plugins for Claude Code, Codex, Gemini, Copilot, OpenCode, OpenClaw and others.
• Local-first storage – memories live in your own SQLite database under your control.

💳 Pricing

Claude-Mem is free and open source under the Apache License 2.0. It runs locally; the only cost is the model usage of the agent you pair it with.

Who Is It For

Developers who live in AI coding agents and are tired of re-explaining their codebase every session — especially on long-running projects where accumulated context is the difference between a helpful agent and a confused one.