About Claworc
Claworc is an open source orchestrator for running a fleet of AI agents from one self hosted dashboard. It works with the OpenClaw, Hermes and NanoClaw agent runtimes and gives each agent its own isolated instance with webchat, a live Chrome browser, a terminal, a file manager and persistent storage that survives restarts and redeployments. Claworc monitors the instances continuously and restarts them automatically if they crash.
It is designed for organizations that want to give every team its own agent, run a shared agent for data analysis, or manage agent bots for clients in one place. Operators can create, start, stop and remove instances, chat with an agent while watching its browser or taking control, stream live logs, and open SSH terminal sessions, including from a local SSH client with a personal key. A skills library packages reusable prompts for one click deployment, with community skills available on Clawhub, and shared folders let instances reuse data. LLM access runs through virtual keys, so real provider credentials stay in the control plane and never enter the containers, and usage statistics are shown by provider or instance. Access is role based, with admins, managers and users who see only their assigned instances, and automated backups support rollbacks.
Claworc is free to download and self host from GitHub; the site publishes no paid plan or hosted pricing. Model usage is paid to whichever LLM providers are connected.
The control plane is a single binary of roughly 20 MB that serves both the dashboard and the proxy layer, and it deploys on Docker for single server setups or on Kubernetes for production scale.