memU
installs/mo · pypistarsforks
How to install memU?
nevamind-ai/memu
$git clone https://github.com/NevaMind-AI/memUAcross Sessions. Across Agents. Across Devices.
README.md
| 1 |  |
| 2 | |
| 3 | <div align="center"> |
| 4 | |
| 5 | # memU |
| 6 | |
| 7 | ### Personal memory, stored as files |
| 8 | |
| 9 | **Across Sessions. Across Agents. Across Devices.** |
| 10 | |
| 11 | [](https://badge.fury.io/py/memu-cli) |
| 12 | [](https://opensource.org/licenses/Apache-2.0) |
| 13 | [](https://www.python.org/downloads/) |
| 14 | [](https://discord.com/invite/hQZntfGsbJ) |
| 15 | [](https://x.com/memU_ai) |
| 16 | |
| 17 | <a href="https://trendshift.io/repositories/17374" target="_blank"><img src="https://trendshift.io/api/badge/repositories/17374" alt="NevaMind-AI%2FmemU | Trendshift" style="width: 250px; height: 55px;" width="250" height="55"/></a> |
| 18 | |
| 19 | </div> |
| 20 | |
| 21 | --- |
| 22 | |
| 23 | memU is a 500-line memory system for AI agents. Agents write what's worth keeping as Markdown; memU stores it, embeds it, and retrieves ranked context in a single call — embeddings are the only model calls it makes. The entire memory logic lives in [`agentic.py`](src/memu/app/agentic.py) + [`service.py`](src/memu/app/service.py); everything else is pluggable storage and embedding transport. |
| 24 | |
| 25 | **Installation is agent-driven.** The guides are written for the agent, not for you. One message is the whole setup — tell your agent: |
| 26 | |
| 27 | > Read https://raw.githubusercontent.com/NevaMind-AI/MemU/main/SKILL.md and follow it to install memU. |
| 28 | |
| 29 | It works for Codex, Claude Code, Cursor, OpenClaw, Hermes, WorkBuddy — and any other agent, via detection. Details in [Host adapters](#host-adapters-memory-for-desktop-coding-agents). |
| 30 | |
| 31 | Want to follow the latest development instead? Install from the newest source (git `main`) — tell your agent: |
| 32 | |
| 33 | > Read https://raw.githubusercontent.com/NevaMind-AI/MemU/main/INSTALL-LATEST.md and follow it to install the latest memU. |
| 34 | |
| 35 | ## Quick start |
| 36 | |
| 37 | ```python |
| 38 | from memu.app import MemoryService |
| 39 | |
| 40 | service = MemoryService( |
| 41 | database_config={"metadata_store": {"provider": "sqlite", "dsn": "sqlite:///memu.sqlite3"}}, |
| 42 | ) |
| 43 | |
| 44 | # 1. Persist agent-prepared memory: recall files (memory/skill tracks) + resources |
| 45 | await service.commit_results( |
| 46 | recall_files=[ |
| 47 | { |
| 48 | "name": "Profile", |
| 49 | "track": "memory", |
| 50 | "description": "who the user is", |
| 51 | "content": "# Profile\n- prefers dark roast coffee\n- ships on Fridays", |
| 52 | }, |
| 53 | { |
| 54 | "name": "deploy-checklist", |
| 55 | "track": "skill", |
| 56 | "description": "how to deploy this repo", |
| 57 | "content": "1. run tests\n2. tag\n3. push", |
| 58 | }, |
| 59 | ], |
| 60 | resource=[{"path": "/abs/path/notes.md", "description": "meeting notes from the launch review"}], |
| 61 | ) |
| 62 | |
| 63 | # 2. See what is stored, across every track |
| 64 | files = await service.list_all_recall_files() |
| 65 | |
| 66 | # 3. Single-shot embedding retrieval over segments / files / resources |
| 67 | context = await service.progressive_retrieve("What should I know about this user's launch preferences?") |
| 68 | ``` |
| 69 | |
| 70 | Or straight from the terminal — no code: |
| 71 | |
| 72 | ```bash |
| 73 | export OPENAI_API_KEY=sk-... # embedding API key — the only model calls memU makes |
| 74 | |
| 75 | npx memu-cli commit results.json # {"recall_files": [...], "resource": [...]} |
| 76 | npx memu-cli list-files |
| 77 | npx memu-cli retrieve "What should I know about this user's launch preferences?" |
| 78 | ``` |
| 79 | |
| 80 | State persists in a local SQLite database (`./data/memu.sqlite3` by default), so commit in one invocation and retrieve in the next. |
| 81 | |
| 82 | ## How it works |
| 83 | |
| 84 |  |
| 85 | |
| 86 | ### The data model |
| 87 | |
| 88 | Memory is a set of **recall files** — one Markdown document per topic (`track="memory"`) or per learned skill (`track="skill"`). Committing a file also writes its search index: |
| 89 | |
| 90 | | Record | What it is | How it's embedded | |
| 91 | |---|---|---| |
| 92 | | **RecallFile** | The Markdown document itself (`na |
People who liked this also viewed
6 picksType
- excel-mcpA Model Context Protocol (MCP) server for working with Excel (.xlsx) files without requiring Microsoft Excel. Enables AI assistants to create, read, modify, and format spreadsheets using openpyxl, fully in Python and cross-platform.MCP ServersOct 2025
- memories-with-lessons-mcp-serverI enhance the existing memory mcp server from the official mcp github, so big thanks and credits for creating thisMCP ServersFeb 2025
- anytype-mcpAn MCP server enabling AI assistants to interact with Anytype - your encrypted, local and collaborative wiki - to organize objects, lists, and more through natural language.MCP ServersJul 2026
- mcp-server-kubernetesMCP Server for kubernetes management commandsMCP ServersJul 2026
- thoughtboxThoughtbox is an intention ledger for agents. Evaluate AI's decisions against its decision-making.MCP ServersJul 2026
- google-drive-mcpA Model Context Protocol (MCP) server that provides secure integration with Google Drive, Docs, Sheets, Slides and Calendar. It allows Claude Desktop and other MCP clients to manage files in Google Drive through a standardized interface.MCP ServersJul 2026
