MicroClaw is a Rust-based agentic AI assistant runtime designed for modern chat surfaces. It provides a channel-agnostic core that enables teams to build one agent and deploy it across multiple chat platforms without duplicating business logic. With approximately 459 GitHub stars and 93 releases, it focuses on practical agentic assistance with persistent memory, scheduled tasks, and cross-platform chat history.
License: MIT
GitHub: microclaw/microclaw
Website: microclaw.ai
- 🦀 Rust-Based Runtime - Single binary deployment with high performance and reliability
- 💬 Multi-Platform - Telegram, Discord, Slack, Feishu/Lark, Matrix, WhatsApp, iMessage, Email, Nostr, Signal, DingTalk, QQ, IRC, and Web UI
- 🧠 Persistent Memory - SQLite with WAL mode, AGENTS.md file-based memory, optional sqlite-vec for semantic memory
- 📅 Scheduled Tasks - Cron-based recurring tasks and one-time scheduled tasks
- 🔧 Agentic Tools - Bash commands, file operations, web search, sub-agent delegation, 20+ built-in tools
- 🔄 Session Resume - Full conversation state (including tool interactions) persisted between messages
- 📝 Context Compaction - Automatic summarization to stay within context limits
- 👥 Sub-Agent - Delegate tasks to parallel agents with restricted tools
- 🔗 MCP Support - Model Context Protocol federation with external tool servers
- ⚡ Skills System - Extensible skill catalog (Anthropic Skills compatible)
- Personal Infra Agent - Run one assistant across your chats with memory, scheduling, and shell/file tooling
- Team Operations Bot - Internal workflows with permission-aware tools, session history, and recurring workflows
- Chat-Based AI Assistant - Lives in your messaging platforms with natural interaction
- Code Development - Find/fix TODOs, search codebases, edit files, manage projects
- System Administration - Run shell commands, analyze logs, monitor systems
- Web Research - Search and fetch web content, summarize information
- Task Automation - Schedule recurring tasks and reminders
- Product Prototyping Runtime - Ship new channels and tools quickly with shared Rust core
| Component |
Technology |
| Language |
Rust (89.3%), TypeScript (7.8%), Shell (1.4%) |
| Database |
SQLite (WAL mode) |
| Optional |
sqlite-vec for embeddings, Docker for sandbox |
| Deployment |
Binary, Docker, Homebrew, One-line installer |
| Architecture |
Channel-agnostic core with platform adapters |
crates/
├── microclaw-core/ # Shared error/types/text modules
├── microclaw-storage/ # SQLite DB + memory domain + usage reporting
├── microclaw-tools/ # Tool runtime primitives + sandbox + helper engines
├── microclaw-channels/ # Channel abstractions and routing boundary
└── microclaw-app/ # App-level support modules (logging, builtin skills, transcribe)
| Component |
Minimum |
Recommended |
| RAM |
512 MB |
2+ GB |
| Disk |
100 MB |
1+ GB |
| Network |
Required for LLM API |
Required |
| Optional |
Docker for sandbox, Node.js for MCP servers |
macOS:
memo (Apple Notes), remindctl (Apple Reminders), icalBuddy + osascript (Apple Calendar)
Linux:
- systemd (for gateway service)
Windows:
- PowerShell, optional pywinauto for desktop automation
MicroClaw supports a wide range of LLM providers:
- Anthropic (native API)
- OpenAI, OpenAI-Codex (OAuth or API key)
- OpenRouter, DeepSeek, Ollama
- Google, Alibaba, Moonshot, Mistral
- Azure, Bedrock, Zhipu, Minimax
- Cohere, Tencent, XAI, HuggingFace, Together
- Custom (manual provider/model/base URL)
- ✅ Open-source and self-hosted
- ✅ MIT License
- ✅ Active development (v0.0.128 - Feb 28, 2026)
- ✅ 459+ GitHub stars, 71 forks, 9 contributors
- ✅ 689 commits, 93 releases
- ✅ Single binary deployment option
- ✅ Comprehensive documentation and generated reference docs
Any questions?
Feel free to contact us. Find all contact information on our contact page.