Verified project record

zeroclaw-labs/zeroclaw

zeroclaw is a self-hosted autonomous AI assistant runtime distributed as a single Rust binary, designed to run on the user's own machine. It connects to roughly 20 LLM providers and reaches more than 30 channels, including Discord, Telegram, Matrix, email, webhooks, and the command line, and acts through tools.

Project overview

zeroclaw pairs provider-agnostic model routing across about 20 providers, where local Ollama or an OpenAI-compatible endpoint can be used in place of remote services, with a unified agent loop that reaches more than 30 channels, and its runtime runs on the user's own machine rather than as a hosted service.

Project type: AI Agent · Infrastructure
Use cases: Automation
Deployment: Refer to project documentation
License: Apache-2.0

Best for

Developers and AI engineers who want an autonomous assistant runtime that runs on their own machine, with their own keys, across many channels and providers.
Users who need to swap or chain model providers, including local Ollama or OpenAI-compatible endpoints, behind one agent.

Key capabilities

Inbound messages from Discord, Telegram, Matrix, email, webhooks, and the CLI are all delivered to the same agent loop.
Configure Anthropic, OpenAI, local Ollama, or any OpenAI-compatible endpoint, with fallback chains and routing.
Gateway and web dashboard for chat, memory browsing, config editing, cron management, and tool inspection.

Limitations and risks

A configured model provider is required for the agent to run; without one, the agent does not operate.
Use requires configuring a provider and TOML config, so it is not usable out of the box without setup.
The runtime executes locally, but inference depends on the configured model provider; using a remote provider such as Anthropic or OpenAI involves an external service, so verify your provider choice against your data boundary requirements. GPU, cost, telemetry, and minimum hardware requirements are not documented.

Getting started

Setup difficulty is documented as easy thanks to a guided installer and a one-shot quickstart command. First success path: curl -fsSL https://raw.githubusercontent.com/zeroclaw-labs/zeroclaw/master/install.sh | sh, then zeroclaw quickstart, then zeroclaw agent -a <alias>.
Before the agent can run, you must configure a model provider and complete the TOML config.

Evidence and sources

README: ZeroClaw is an agent runtime — a single Rust binary you configure and run. It talks to LLM providers (Anthropic, OpenAI, Ollama, and ~20 others), reaches the world through 30+ cha…
Release: v0.8.6
README: Inbound messages from Discord, Telegram, Matrix, email, webhooks, CLI — all delivered to the same agent loop.
README: Configure Anthropic, OpenAI, local Ollama, or any OpenAI-compatible endpoint. Fallback chains and routing keep the agent running when a provider flakes.
README: HTTP / WebSocket gateway for clients, with a web dashboard for chat, memory browsing, config editing, cron management, and tool inspection.

Source updated at: 2026-10-19T18:24:14Z

Open original GitHub project