# MemohAI - The Agent OS

Memoh is an Agent OS built for hackers and AI developers. It is self-hosted, highly concurrent, and has zero telemetry. It features containerized isolation and long-term memory for secure agent orchestration.

## Core Capabilities
- **Multi-Bot & Multi-User**: Deploy squads of agents that recognize users in group chats.
- **Containerized Isolation**: Every agent gets its own sandbox via containerd, ensuring the host machine remains safe.
- **Memory Engineering**: True long-term memory with hybrid search (Dense + Sparse + BM25) and LLM-driven fact extraction.
- **Omnichannel Native**: Direct hooks into Telegram, Discord, Matrix, Lark, WeChat, and Email.
- **MCP Ecosystem**: Native Model Context Protocol (MCP) support for local machine and external API control.
- **Browser Automation**: Headless browser automation backed by Playwright.
- **Skills & Subagents**: Modular design allowing agents to use skills or spawn sub-agents.
- **Workflow Automation**: Built-in cron jobs for scheduled tasks.
- **Access Control (ACL)**: Tight ACLs based on user IDs or channels.
- **Multi-Model Routing**: Support for OpenAI, Anthropic, or local Ollama nodes.
- **Instant Deployment**: Deploy the entire cluster (containers, memory stores, models) in under three minutes via a single Docker script.

## Quick Start
```bash
curl -sL https://memoh.sh | bash
```

## Links
- Website: https://memoh.ai
- Waitlist: https://memoh.ai/waitlist
- Blog: https://memoh.ai/blogs
- GitHub: https://github.com/felinics/Memoh
- Plugin Market: https://github.com/felinics/supermarket

## Plain-Text Indexes
- Help Center (plans, billing, desktop app, privacy): https://memoh.ai/help/llms.txt
- Documentation (setup and feature references): https://docs.memoh.ai/llms.txt

## Blog Posts
- Introducing Memoh: A Cloud Computer for Every Agent: https://memoh.ai/blogs/cloud-computer-for-every-agent
- Desktop workspace display: https://memoh.ai/blogs/workspace-desktop
- Discuss mode for proactive group chat agents: https://memoh.ai/blogs/discuss-mode
- Introduction to Memoh: https://memoh.ai/blogs/introduction-to-memoh