NevaMind-AI/memU

Personal memory across agents

What it solves

memU provides a shared, persistent memory system for AI agents. It solves the problem of "forgetting" across different sessions, different AI agents (like Cursor, Claude Code, or ChatGPT), and different devices, by storing user-specific knowledge and reusable skills in a centralized "wiki" format.

How it works

memU acts as a sidecar to desktop coding agents. It uses host adapters to monitor session logs and a two-part process to manage memory:

  1. Memorization: A background task slices session history into jobs. An agent then analyzes these logs to decide whether to create a new "skill" (a Markdown file describing a workflow) or update an existing one. These skills are then embedded and stored in a database.
  2. Retrieval: memU patches the agent's instruction files to ensure the agent runs a retrieval command before answering a prompt, bringing relevant stored skills and memories into the current context.

The system can be hosted in the cloud (memu.so) or self-hosted locally using SQLite or Postgres.

Who it’s for

It is designed for developers using multiple AI coding assistants (such as Cursor, Claude Code, Codex, and others) who want their agents to learn from their past interactions and maintain a consistent knowledge base across tools.

Highlights

  • Cross-Agent Compatibility: Works across various hosts including Cursor, Claude Code, ChatGPT (Work mode), and OpenClaw.
  • Automatic Skill Extraction: Automatically turns agent history into reusable Markdown-based workflows.
  • Lightweight Core: The core memory logic is approximately 500 lines of code.
  • Flexible Storage: Supports local SQLite, Postgres (via pgvector), and a managed cloud backend.

Related

  • Project
    TencentCloud/TencentDB-Agent-Memory

    A centralized memory server and management hub that allows AI agents to share and reuse distilled experience, skills, and project knowledge across sessions and frameworks.

  • Project
    mem9-ai/mem9

    A persistent memory layer for AI agents that enables shared context and recall across sessions, machines, and different agent runtimes.

  • Project
    akitaonrails/ai-memory

    ai‑memory is a Rust‑based server that records every observation a coding‑assistant makes (prompts, tool calls, session boundaries) as plain markdown files in a git‑backed wiki. The server can be run locally, in Docker/Podman, or on a homelab box and is designed to be shared across many agents (Claude Code, Codex, Cursor, Gemini CLI, etc.), machines and team members. Hooks posted by agents are consolidated into readable wiki pages at session end; the next session receives a concise handoff so it can continue exactly where the previous one stopped. Storage is human‑readable markdown plus a SQLite index (FTS5, entity and optional vector search). By default it needs no LLM or embedding keys; optional Anthropic/OpenAI integrations add LLM‑generated summaries and semantic search. Multi‑user authentication, audit logs and per‑person attribution are built in. Install via Arch AUR, Docker wrapper, or native binary, start the server, register a client and its hooks, and the system works transparently.

  • Project
    zilliztech/memsearch

    memsearch is a Python library and CLI that records AI coding‑assistant chats as Markdown, builds a hybrid (BM25 + dense) Milvus index, and provides cross‑platform recall via plugins for Claude Code, Codex, DeepSeek Harness, OpenClaw, and OpenCode. It supports local ONNX embeddings, optional Zilliz Cloud backend, background project/user notes, and automatic skill distillation.

  • Project
    mnemon-dev/mnemon

    A persistent, LLM-supervised memory layer for AI agents that enables cross-session knowledge retention across multiple AI frameworks using a four-graph knowledge store.