microclaw/microclaw

🦀An agentic AI assistant that lives in your chats, inspired by nanoclaw and incorporating some of its design ideas. Built with Rust 🦀

📚 What is MicroClaw?

MicroClaw is a self‑hosted, Rust‑based AI‑agent platform that can run either as a long‑living server, as a native desktop “Work” app, or be embedded in any Rust program via its microclaw-sdk. All three entry points share the same core Agent Engine – the part that talks to LLM providers, runs tools, keeps memory, and enforces safety policies.

🎯 Who is it for?

Use‑case Recommended surface
Always‑on chat bots, web/API endpoints, scheduled automation MicroClaw Server (macOS, Linux, Windows)
Local, workspace‑centric “co‑worker” that lives on your desktop MicroClaw Work – a native GPUI app (Apple Silicon macOS 13+, Linux/Windows previews)
Adding an AI‑agent to an existing Rust codebase microclaw‑sdk – embed the engine directly

⚙️ Core capabilities

Area What you get
Agent execution Multi‑step tool use, parallel read‑only waves, planning, sub‑agents, completion contracts
Continuity Resumable sessions, checkpoints, durable outbound delivery, scheduling, cancellation
Memory & learning File‑based + SQLite memory, semantic recall, temporal knowledge graph, governed skill evolution
Extensibility Skills, plugins, hooks, MCP, ClawHub, A2A, ACP – all plug into the same runtime
Interfaces Rust SDK, native Work UI, local Web UI, HTTP/SSE/WebSocket APIs, dozens of chat adapters (Telegram, Discord, Slack, Matrix, etc.)
Safety & ops Tool‑approval gates, scoped capability grants, Docker sandboxing, egress policy, secret redaction, audit trails

🛠️ How to get started

  1. Work (desktop) – brew tap microclaw/tap && brew install --cask microclaw-work (Apple Silicon macOS) or download the Linux/Windows preview from the v0.9.0 release.
  2. Server – one‑liner installer: curl -fsSL https://microclaw.org/install.sh | bash (macOS/Linux) or the PowerShell equivalent for Windows. Then run microclaw doctor && microclaw setup && microclaw start and open http://127.0.0.1:10961.
  3. SDK – add to Cargo.toml:
    microclaw-sdk = { version = "0.9.0", features = ["full"] }
    and follow the quick‑start code snippet to spin up an agent programmatically.

📖 Documentation & community

  • Docs hub – docs/README.md links to getting‑started guides, SDK concepts, tool catalog, security model, and operations runbooks.
  • Support – Discord, Reddit, and a public issue tracker. Security disclosures go through SECURITY.md.
  • Release cadence – stable branch for production, main for rapid development. Latest stable is v0.9.0.

🤔 Why choose MicroClaw?

  • Single engine, multiple front‑ends – you can start with the desktop app, later switch to a server, or embed the same engine without rewriting logic.
  • Rust‑first – high performance, low‑footprint, and easy to ship as a single binary; SQLite persistence keeps dependencies minimal.
  • Durable, resumable work – the runtime checkpoints after each tool call, so long‑running or interrupted tasks can be recovered safely.
  • Provider‑agnostic – native Anthropic support plus any OpenAI‑compatible or local model via a unified message format.
  • Safety baked in – built‑in guardrails, sandboxed tool execution, and audit trails make it suitable for production workloads.

📦 Quick tech snapshot

  • Language: Rust (requires 1.93+, built on 1.96 toolchain for v0.9.0)
  • License: MIT
  • Packaging: Homebrew, Docker (ghcr.io/microclaw/microclaw:latest), source (cargo build --release)
  • Platform support: macOS (Apple Silicon), Linux (x86_64/arm64), Windows (portable preview)

MicroClaw lets you run sophisticated, stateful AI agents locally or as a service, with a focus on reliability, extensibility, and safety—all from a single Rust codebase.

Related

  • Project
    qhkm/zeptoclaw

    A fast, secure, and local-first personal AI assistant infrastructure written in Rust, providing sandboxed autonomy and multi-provider LLM support in a single 6MB binary.

  • Project
    clawdotnet/openclaw.net

    OpenClaw.NET is a .NET‑native, self‑hosted AI‑agent runtime with a desktop Avalonia UI (AgentQi Companion). It runs a gateway that talks to any OpenAI‑compatible model, offers streaming, tool execution, session memory, and many integrations (REST, WebSocket, MCP, 80+ tools, channel adapters). The gateway can be compiled with NativeAOT, making it suitable for edge or on‑prem deployments. Users can start via a ready‑made desktop bundle or from source, configure the assistant through chat commands, and extend functionality with .NET plugins or reusable MetaSkill DAG workflows.

  • Project
    zeroclaw-labs/zeroclaw

    A self-hosted agent runtime written in Rust that connects LLM providers to multiple communication channels and hardware tools while ensuring user ownership of data and infrastructure.

  • Project
    ultraworkers/claw-code

    Claw Code is a Rust‑based CLI harness (`claw`) for running autonomous LLM agents. It supports Anthropic, OpenAI, and local providers, offers health checks, interactive and one‑off prompting, and includes extensive docs for building, configuring, and testing. Build from source (or install the upstream `agent-code` binary) and use it to manage agent‑driven workflows.

  • Project
    johnson7788/MultiUserClaw

    MultiUserClaw is a Docker‑based SaaS framework that lets you run isolated AI assistants (Hermes agents) for many users. It includes a React front‑end, a FastAPI gateway for auth, container management and LLM proxying, per‑user Hermes containers, PostgreSQL storage, and features like knowledge bases, skill store, cron jobs, multi‑channel bots, and multi‑model support.