Red Hat Engineer Develops Tank OS, a Tool that Makes OpenClaw Safer for Enterprises
In Focus
- OpenClaw installs AI agents on local computers
- Task OS supports safe deployment of OpenClaw AI agents in enterprise setups
- The open-source tool launches OpenClaw AI agents when a computer starts
Red Hat engineer Sally O’Malley has released an open-source tool called Tank OS. The new tool supports safe deployment and management of OpenClaw agents. Tank OS makes OpenClaw easier to maintain at scale and is designed for IT professionals who manage multiple agents in corporate setups. Users who deploy OpenClaw AI agents on their computers will also find the tool useful.
“This was a fun project that I put together on the weekend that I knew would be a really good fit for AI and where we’re going,” O’Malley informed TechCrunch.
O’Malley Works With OpenClaw’s Founder
OpenClaw is an open-source project that installs an AI agent on a local computer. Numerous startups and companies have started creating better ways to work with the AI-powered tool. Companies are also developing safer alternatives to OpenClaw.
Early this month, reports that Microsoft was developing an OpenClaw-like agent system to enhance its Office 365 Copilot emerged. But what makes the Red Hat enterprise tool stand out is the fact that its developer is OpenClaw’s maintainer.
O’Malley is among the few engineers who works with OpenClaw creator Peter Steinberger, whom OpenAI recently hired, to determine which bugs get fixed and what features are developed. She works on ensuring that OpenClaw supports enterprise use cases and with Red Hat’s Linux operating system.
Why Did the Red Hat Engineer Launch Tank OS
O’Malley’s decision to develop Tank OS was the thought of what will happen when OpenClaw gains popularity among enterprises. She started with an open-source container tool known as Podman, which a colleague developed.
Containers run apps separately from an underlying computer. For instance, they can support a Linux app on a Mac or Windows computer. Tank OS installs OpenClaw onto Red Hat’s Fedora Linux OS in a Podman container. The open-source tool supports the OpenClaw container security in a Podman by launching the AI agent when a computer starts.
Red Hat’s Tank OS contains the features required to ensure OpenClaw can run without human supervision, including the ability to store API keys. Users can operate multiple Tank OS instances to complete different tasks in a single machine.
What Security Risks Do AI Agents Face?
There are instances where AI agents have performed unintended actions like deleting emails or exposing messages. Increasingly, hackers are targeting OpenClaw users with malware attacks. Although the OpenClaw team is working on making the AI agent safer, O’Malley notes that the new tool is ‘incredibly powerful’ and can pose risks if not configured properly. The tool is not designed for beginners. Users should be comfortable installing and managing software to use it.
