[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-74637":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":10,"language":11,"languages":10,"totalLinesOfCode":10,"stars":12,"forks":13,"watchers":14,"openIssues":15,"contributorsCount":16,"subscribersCount":16,"size":16,"stars1d":17,"stars7d":18,"stars30d":19,"stars90d":16,"forks30d":16,"starsTrendScore":20,"compositeScore":21,"rankGlobal":10,"rankLanguage":10,"license":22,"archived":23,"fork":23,"defaultBranch":24,"hasWiki":23,"hasPages":25,"topics":26,"createdAt":10,"pushedAt":10,"updatedAt":33,"readmeContent":34,"aiSummary":35,"trendingCount":16,"starSnapshotCount":16,"syncStatus":36,"lastSyncTime":37,"discoverSource":38},74637,"Memoh","memohai\u002FMemoh","memohai","✨ The open-source multi-agent platform. Every agent gets its own computer, desktop, network, and long-term memory. ","https:\u002F\u002Fdocs.memoh.ai",null,"Go",1895,176,12,9,0,63,109,273,189,19.74,"GNU Affero General Public License v3.0",false,"main",true,[27,28,29,30,31,32],"agent","ai","ai-companion","ai-memory","openclaw","personal-assistant","2026-06-12 02:03:26","\u003Cdiv align=\"right\">\n  \u003Cspan>[\u003Ca href=\".\u002FREADME.md\">English\u003C\u002Fa>]\u003Cspan>\n  \u003C\u002Fspan>[\u003Ca href=\".\u002FREADME_CN.md\">简体中文\u003C\u002Fa>]\u003C\u002Fspan>\n\u003C\u002Fdiv>  \n\n\u003Cdiv align=\"center\">\n  \u003Cimg src=\".\u002Fassets\u002Flogo.png\" alt=\"Memoh\" height=\"80\">\n  \u003Ch1>Memoh\u003C\u002Fh1>\n  \u003Cp>Self hosted, always-on AI agent orchestrator in containers.\u003C\u002Fp>\n  \u003Cdiv align=\"center\">\n    \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fpackage-json\u002Fv\u002Fmemohai\u002FMemoh\" alt=\"Version\" \u002F>\n    \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Flicense\u002Fmemohai\u002FMemoh\" alt=\"License\" \u002F>\n    \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fmemohai\u002FMemoh?style=social\" alt=\"Stars\" \u002F>\n    \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fforks\u002Fmemohai\u002FMemoh?style=social\" alt=\"Forks\" \u002F>\n    \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Flast-commit\u002Fmemohai\u002FMemoh\" alt=\"Last Commit\" \u002F>\n    \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fissues\u002Fmemohai\u002FMemoh\" alt=\"Issues\" \u002F>\n    \u003Ca href=\"https:\u002F\u002Fdeepwiki.com\u002Fmemohai\u002FMemoh\">\n      \u003Cimg src=\"https:\u002F\u002Fdeepwiki.com\u002Fbadge.svg\" alt=\"DeepWiki\" \u002F>\n    \u003C\u002Fa>\n    \u003Ca href=\"https:\u002F\u002Ft.me\u002Fmemohai\">\n      \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FTelegram-Group-26A5E4?logo=telegram&logoColor=white\" alt=\"Telegram\" \u002F>\n    \u003C\u002Fa>\n    \u003Ca href=\"https:\u002F\u002Fdocs.memoh.ai\">\n      \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FDocs-memoh.ai-3eaf7c?logo=readthedocs&logoColor=white\" alt=\"Documentation\" \u002F>\n    \u003C\u002Fa>\n  \u003C\u002Fdiv>\n\u003C\u002Fdiv>\n\n**Memoh(\u002Fˈmemoʊ\u002F)** is an always-on, containerized AI agent orchestrator. Create multiple AI bots, each running in its own isolated container with persistent memory, and interact with them across Telegram, Discord and so on. Bots can execute commands, edit files, browse the web, call external tools via MCP, and remember everything — like giving each bot its own computer and brain.\n\n## Quick Start\n\nMemoh is distributed in two forms:\n\n### ⚙️ Deploy Version\n\nThe self-hosted server stack for always-on, multi-user or multi-tenant usage. Use this when you want Memoh running on a server, VM, NAS, or Kubernetes cluster, with bots available through Web UI and external channels such as Telegram, Discord, Lark, WeChat, Email, and more.\n\n\u003Cdetails>\n\u003Csummary>\u003Cstrong>🐳 Deploy Memoh Server\u003C\u002Fstrong>\u003C\u002Fsummary>\n\nUse the deploy version when Memoh should be reachable by multiple users, run bots continuously, or connect to public\u002Fprivate messaging channels. The default Docker deployment starts the server, Web UI, database migrations, container runtime support, and the services needed for workspace containers.\n\nOne-click install (**requires [Docker](https:\u002F\u002Fwww.docker.com\u002Fget-started\u002F)**):\n\n```bash\ncurl -fsSL https:\u002F\u002Fmemoh.sh | sh\n```\n\n*Silent install with all defaults: `curl -fsSL ... | sh -s -- -y`*\n\nOr manually:\n\n```bash\ngit clone --depth 1 https:\u002F\u002Fgithub.com\u002Fmemohai\u002FMemoh.git\ncd Memoh\ncp conf\u002Fapp.docker.toml config.toml\n# Edit config.toml\ndocker compose up -d\n```\n\nKubernetes deployment (**uses Kubernetes workspaces by default**):\n\n```bash\nkubectl apply -k deploy\u002Fkubernetes\n```\n\n> **Install a specific version:**\n> ```bash\n> curl -fsSL https:\u002F\u002Fmemoh.sh | MEMOH_VERSION=v0.6.0 sh\n> ```\n>\n> **Use CN mirror for slow image pulls:**\n> ```bash\n> curl -fsSL https:\u002F\u002Fmemoh.sh | USE_CN_MIRROR=true sh\n> ```\n>\n> Do not run the whole installer with `sudo`. The installer will use `sudo docker`\n> internally if Docker requires it. On macOS or if your user is in the `docker`\n> group, `sudo` is not required for Docker either.\n\nVisit \u003Chttp:\u002F\u002Flocalhost:8082> after startup. Default login: `admin` \u002F `admin123`.\n\nSee [DEPLOYMENT.md](DEPLOYMENT.md) for custom configuration and production setup.\n\n\u003C\u002Fdetails>\n\n### 🖥️ Desktop Version\n\nThe native client for personal\u002Flocal use. It starts and manages a local Memoh server on your computer, bundles the CLI and local runtime resources, and is the easiest way to try Memoh without maintaining a separate server deployment.\n\n\u003Cdetails>\n\u003Csummary>\u003Cstrong>⏬ Install Memoh Desktop\u003C\u002Fstrong>\u003C\u002Fsummary>\n\nUse the desktop version when you want a local app that manages Memoh for you. It is designed for single-user desktop workflows, local memory, and local\u002FDocker-backed workspaces. The desktop app runs its own local server, so it is separate from the deploy version above.\n\n1. Download the installer for your platform from the [Memoh Desktop download page](https:\u002F\u002Fmemoh.ai\u002Fdesktop).\n2. Open Memoh. The app starts the local server, prepares local storage, and connects the UI automatically.\n3. Optional: install the bundled `memoh` CLI from the app menu if you want terminal access to the same local server.\n\nChoose the deploy version instead if you need a shared server, remote access, production uptime, or channel integrations that should keep running while your desktop is offline.\n\n\u003C\u002Fdetails>\n\nDocumentation entry points:\n\n- [About Memoh](https:\u002F\u002Fdocs.memoh.ai\u002Fabout)\n- [Providers & Models](https:\u002F\u002Fdocs.memoh.ai\u002Fgetting-started\u002Fprovider-and-model)\n- [Bot Setup](https:\u002F\u002Fdocs.memoh.ai\u002Fgetting-started\u002Fbot)\n- [Sessions & Discuss Mode](https:\u002F\u002Fdocs.memoh.ai\u002Fgetting-started\u002Fsessions)\n- [Channels](https:\u002F\u002Fdocs.memoh.ai\u002Fgetting-started\u002Fchannels)\n- [Skills](https:\u002F\u002Fdocs.memoh.ai\u002Fgetting-started\u002Fskills)\n- [Supermarket](https:\u002F\u002Fdocs.memoh.ai\u002Fgetting-started\u002Fsupermarket)\n- [Slash Commands](https:\u002F\u002Fdocs.memoh.ai\u002Fgetting-started\u002Fslash-commands)\n\n## Why Memoh?\n\nMemoh is built for **always-on continuity** — an AI that stays online, and a memory that stays yours.\n\n- **Lightweight & Fast**: Built with Go as home\u002Fstudio infrastructure, runs efficiently on edge devices.\n- **Containerized by default**: Each bot gets an isolated container with its own filesystem, network, and tools.\n- **Hybrid split**: Cloud inference for frontier model capability, local-first memory and indexing for privacy.\n- **Multi-user first**: Explicit sharing and privacy boundaries across users and bots.\n- **Full graphical configuration**: Configure bots, channels, MCP, skills, and all settings through a modern web UI — no coding required.\n\n## Features\n\n### Core\n\n- 🤖 **Multi-Bot & Multi-User**: Create multiple bots that chat privately, in groups, or with each other. Bots distinguish individual users in group chats and remember each person's context.\n- 📦 **Containerized Workspaces**: Each bot can run in its own isolated workspace container with a dedicated filesystem, network, tools, snapshots, data export\u002Fimport, and versioning.\n- 🖥️ **Desktop Environment in Containers**: Give a bot a full graphical desktop inside its workspace container, including VNC access and a headed browser for sites and workflows that need a real GUI session.\n- 🗂️ **Persistent File System**: Every bot has a writable home directory that survives restarts, upgrades, and migrations. Bots can read, write, and organize files freely; you can browse, upload, download, and edit them visually through the web UI's file manager.\n- 🧠 **Memory Engineering**: LLM-driven fact extraction, hybrid retrieval (dense + sparse + BM25), provider-based long-term memory, memory compaction, and separate session-level context compaction. Pluggable backends: Built-in (off \u002F sparse \u002F dense), [Mem0](https:\u002F\u002Fmem0.ai), OpenViking.\n- 💬 **Broad Channel Coverage**: Telegram, Discord, Lark (Feishu), QQ, Matrix, Misskey, DingTalk, WeCom, WeChat, WeChat Official Account, Email (Mailgun \u002F SMTP \u002F Gmail OAuth), and built-in Web UI.\n\n### Agent Capabilities\n\n- 🔧 **MCP (Model Context Protocol)**: Full MCP support (HTTP \u002F SSE \u002F Stdio \u002F OAuth). Connect external tool servers for extensibility; each bot manages its own independent MCP connections.\n- 🌐 **Browser Use**: Drive Chromium\u002FFirefox through Playwright for navigation, form filling, screenshots, accessibility trees, and tab control. When headless mode is not enough, run a headed browser in the workspace desktop.\n- 🖱️ **Computer Use**: Observe and operate the bot's workspace desktop through visual state and input events, including clicking, typing, scrolling, and recovering from GUI flows that cannot be handled headlessly.\n- 🎭 **Skills, Supermarket & Subagents**: Define bot behavior through modular skills, install curated skills and MCP templates from Supermarket, and delegate complex tasks to sub-agents with independent context.\n- 💭 **Sessions & Discuss Mode**: Use chat, discuss, schedule, heartbeat, and subagent sessions with slash-command control and session status inspection.\n- ⏰ **Automation**: Cron-based scheduled tasks and periodic heartbeat for autonomous bot activity.\n\n### Management\n\n- 🖥️ **Web UI**: Modern dashboard (Vue 3 + Tailwind CSS) — streaming chat, tool call visualization, file manager, visual configuration for all settings. Dark\u002Flight theme, i18n.\n- 💻 **Desktop App**: Native Memoh client for personal\u002Flocal use, with a self-managed local server, embedded Qdrant, bundled CLI, local workspace support, and system tray lifecycle controls.\n- 🔐 **Access Control**: Priority-based ACL rules with presets, allow\u002Fdeny effects, and scope by channel identity, channel type, or conversation.\n- 🧪 **Multi-Model**: OpenAI-compatible, Anthropic, Google, OpenAI Codex, GitHub Copilot, and Edge TTS providers. Per-bot model assignment, provider OAuth, and automatic model import.\n- 🎙️ **Speech & Transcription**: Bots can speak through 10+ TTS providers (Edge, OpenAI, ElevenLabs, Deepgram, Azure, Google, MiniMax, Volcengine, Alibaba, OpenRouter) and listen — voice messages received from Telegram, Discord, etc. are auto-transcribed via STT models (OpenAI \u002F OpenRouter), and bots can transcribe any audio file on demand through a built-in tool.\n- 🚀 **Server Deploy**: Docker Compose and Kubernetes deployment paths for always-on server usage, with automatic migration, container runtime setup, and supporting services for workspace containers.\n\n## Memory System\n\nMemoh ships with a **fully self-hosted memory engine** out of the box — no external API, no SaaS dependency. Every bot remembers what you've told it across sessions, days, and platforms; in group chats, each user's memories are kept separately so the bot doesn't mix you up with the rest.\n\n### Built-in Memory (default)\n\nThree modes, switchable per bot from the web UI:\n\n| Mode | Backend | When to use |\n|------|---------|-------------|\n| **Off** | Plain file storage, no vector search | Small bots, debugging, or when you want minimal moving parts |\n| **Sparse** | Neural sparse vectors via a local model + BM25 | Zero API cost, runs entirely on your machine, strong recall for short factual memories |\n| **Dense** | Embedding model + Qdrant vector DB | Best semantic recall — finds memories by meaning, not just keywords |\n\nUnder the hood:\n\n- **LLM-driven fact extraction** — every conversation turn is parsed, deduplicated, and stored as structured memories rather than raw transcripts.\n- **Hybrid retrieval** — dense vectors, sparse vectors and BM25 are combined and re-ranked, so both \"what was that API key\" (lexical) and \"the project I told you about last week\" (semantic) hit reliably.\n- **Memory compaction** — redundant or stale entries are periodically merged by an LLM, keeping the index small and recall sharp.\n- **Inspect & edit anything** — browse, search, manually create\u002Fedit memories, rebuild the whole index, and visualize the vector manifold (Top-K distribution & CDF curves) from the web UI.\n\n### Other providers\n\nIf you'd rather plug into an existing memory service, Memoh also supports [**Mem0**](https:\u002F\u002Fmem0.ai) (SaaS) and **OpenViking** (self-hosted or SaaS) as drop-in alternatives — same bot binding, same chat experience, just a different backend.\n\nSee the [documentation](https:\u002F\u002Fdocs.memoh.ai\u002Fmemory-providers\u002F) for full setup details.\n\n## Gallery\n\n\u003Ctable>\n  \u003Ctr>\n    \u003Ctd>\u003Cimg src=\".\u002Fassets\u002Fgallery\u002F01.png\" alt=\"Gallery 1\" width=\"100%\">\u003C\u002Ftd>\n    \u003Ctd>\u003Cimg src=\".\u002Fassets\u002Fgallery\u002F02.png\" alt=\"Gallery 2\" width=\"100%\">\u003C\u002Ftd>\n  \u003C\u002Ftr>\n  \u003Ctr>\n    \u003Ctd>\u003Cstrong text-align=\"center\">Chat\u003C\u002Fstrong>\u003C\u002Ftd>\n    \u003Ctd>\u003Cstrong text-align=\"center\">Container\u003C\u002Fstrong>\u003C\u002Ftd>\n  \u003C\u002Ftr>\n  \u003Ctr>\n    \u003Ctd>\u003Cimg src=\".\u002Fassets\u002Fgallery\u002F03.png\" alt=\"Gallery 3\" width=\"100%\">\u003C\u002Ftd>\n    \u003Ctd>\u003Cimg src=\".\u002Fassets\u002Fgallery\u002F04.png\" alt=\"Gallery 4\" width=\"100%\">\u003C\u002Ftd>\n  \u003C\u002Ftr>\n  \u003Ctr>\n    \u003Ctd>\u003Cstrong text-align=\"center\">Providers\u003C\u002Fstrong>\u003C\u002Ftd>\n    \u003Ctd>\u003Cstrong text-align=\"center\">File Manager\u003C\u002Fstrong>\u003C\u002Ftd>\n  \u003C\u002Ftr>\n  \u003Ctr>\n    \u003Ctd>\u003Cimg src=\".\u002Fassets\u002Fgallery\u002F05.png\" alt=\"Gallery 5\" width=\"100%\">\u003C\u002Ftd>\n    \u003Ctd>\u003Cimg src=\".\u002Fassets\u002Fgallery\u002F06.png\" alt=\"Gallery 6\" width=\"100%\">\u003C\u002Ftd>\n  \u003C\u002Ftr>\n  \u003Ctr>\n    \u003Ctd>\u003Cstrong text-align=\"center\">Scheduled Tasks\u003C\u002Fstrong>\u003C\u002Ftd>\n    \u003Ctd>\u003Cstrong text-align=\"center\">Token Usage\u003C\u002Fstrong>\u003C\u002Ftd>\n  \u003C\u002Ftr>\n\u003C\u002Ftable>\n\n\n## Sub-projects Born for This Project\n\n- [**Twilight AI**](https:\u002F\u002Fgithub.com\u002Fmemohai\u002Ftwilight-ai) — A lightweight, idiomatic AI SDK for Go — inspired by [Vercel AI SDK](https:\u002F\u002Fsdk.vercel.ai\u002F). Provider-agnostic (OpenAI, Anthropic, Google), with first-class streaming, tool calling, MCP support, and embeddings.\n\n## Star History\n\n[![Star History Chart](https:\u002F\u002Fapi.star-history.com\u002Fsvg?repos=memohai\u002FMemoh&type=date&legend=top-left)](https:\u002F\u002Fwww.star-history.com\u002F#memohai\u002FMemoh&type=date&legend=top-left)\n\n## Contributors\n\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fmemohai\u002FMemoh\u002Fgraphs\u002Fcontributors\">\n  \u003Cimg src=\"https:\u002F\u002Fcontrib.rocks\u002Fimage?repo=memohai\u002FMemoh\" \u002F>\n\u003C\u002Fa>\n\n## Community\n\n- 🌐 [**Website**](https:\u002F\u002Fmemoh.ai)\n- 📚 [**Documentation**](https:\u002F\u002Fdocs.memoh.ai) — setup, concepts, and guides\n- 🤝 [**Cooperation**](mailto:business@memoh.net) — business@memoh.net\n- 💬 [**Telegram Group**](https:\u002F\u002Ft.me\u002Fmemohai) — community chat & support\n- 🛒 [**Supermarket**](https:\u002F\u002Fgithub.com\u002Fmemohai\u002Fsupermarket) — curated skills & MCP templates\n\n---\n\n**LICENSE**: AGPLv3\n\nMade with ❤️ by MemohAI Team,\n\nCopyright (C) 2026 MemohAI (memoh.ai). All rights reserved.\n","Memoh 是一个自托管、始终在线的AI代理平台，运行于容器中。它支持创建多个具有长期记忆功能的AI机器人，并能将这些机器人连接到Telegram、Discord、飞书等通讯工具上。核心功能包括但不限于执行命令、编辑文件、网页浏览及调用外部工具等，所有操作都通过MCP（Memoh Command Protocol）实现。每个AI代理都在独立的容器内运行，确保了数据隔离与安全性。此项目特别适合需要个性化AI助手的企业或个人开发者使用，无论是作为智能客服还是私人助理都非常合适。开发语言为Go，采用AGPLv3许可证开放源代码。",2,"2026-06-11 03:50:12","high_star"]