memory systems for AI agents

The best-benchmarked open-source AI memory system. Spatial memory model, automatic organization, cross-session recall.

A high-throughput and memory-efficient inference and serving engine for LLMs.

A Claude Code plugin that automatically captures everything Claude does during your coding sessions, compresses it with AI, and injects relevant context back.

A single CLAUDE.md file to improve Claude Code behavior, derived from Andrej Karpathy's observations on LLM coding pitfalls.

All-in-One RAG Framework. Multimodal document parsing with knowledge graph integration for intelligent retrieval.

Fast and accurate AI-powered file content types detection. Understands any file format for intelligent processing.

Desktop app to manage markdown knowledge bases. Visual graph, search, and organize your AI brain files.

Train any agent simply by talking. Reinforcement learning from conversation for persistent skill acquisition.

Build your own AI SRE agents. The open source toolkit for the AI era with persistent observability memory.

Procedural memory for AI coding agents: transforms scattered session history into persistent, cross-agent memory so every agent learns from every other.

Graph-first cognitive daemon v8.2 for MCP-enabled dev environments. Persistent knowledge graph, VS Code extension, dashboard, dream cycles, multi-repo support. The graph becomes the source of truth.

Open LLM Knowledge Base. Build, query, and manage AI knowledge bases with vector search and structured retrieval.

Shared memory for your team's coding agents. Captures every agent session and turns it into a shared, evolving asset with a Karpathy-style wiki. 49% speedup.

The open-source memory operating system for AI agents. Persistent memory, semantic search, loop detection, agent messaging, crash recovery, and real-time observability.

A personal context store for AI agents and assistants. Reuse your existing coding agent CLI with built-in skills and tools to capture, search, and reuse project context.

The missing memory layer for coding agents. Trace, capture, and replay agent sessions for debugging and knowledge transfer.

One brain, many harnesses. Portable .agent/ folder (memory + skills + protocols) that plugs into Claude Code, Cursor, Windsurf, OpenCode, OpenClaw, Hermes, or DIY Python.

A workflow-driven AI agent framework that executes YAML-defined decision trees. Bridging rigid automation and unpredictable agents with persistent state.

Claude Code operating system with built-in memory management, knowledge bases, and cross-session persistence.

Self-evolving skill engine — agents learn and share skills across sessions, 46% fewer tokens through learned patterns.

Persistent memory for AI agents. Context window management, semantic search across memories, and automatic memory consolidation.

Production-tested persistent knowledge base pattern. Karpathy-inspired wiki that self-heals and compounds over time.

Personal knowledge brain architecture. Multi-source ingestion, Obsidian integration, agent-queryable knowledge graph.