Quick Summary
Over 230 malicious packages disguised as legitimate tools have been published for the OpenClaw AI assistant, previously known as Moltbot. These packages, called skills, are designed to steal sensitive data such as API keys, wallet private keys, and browser passwords. Security researchers have identified that the malware can bypass security measures on macOS and Windows, posing significant risks to users’ data and privacy.
The Securityish Brief
In a concerning development, more than 230 malicious packages, referred to as skills, have been published for the OpenClaw AI assistant between January 27th and February 1st. These packages masquerade as legitimate utilities, including cryptocurrency trading tools and social media services, while actually delivering malware that steals sensitive information. The skills were identified on ClawHub, the official registry for OpenClaw, and GitHub, indicating a widespread threat to users.
The malicious skills are linked to a single developer and are primarily clones with randomized names, some of which have been downloaded thousands of times. Each skill includes extensive documentation to appear credible, often mentioning a tool called ‘AuthTool’ that is, in reality, a malware delivery mechanism. This mechanism operates differently on macOS and Windows, with macOS versions using base64-encoded commands to download malware.
The malware, identified as a variant of NovaStealer, targets critical data such as cryptocurrency exchange API keys, wallet files, browser passwords, and SSH keys. A report from Koi Security found 341 malicious skills on ClawHub, revealing the scale of this ongoing campaign.
Peter Steinberger, the creator of OpenClaw, acknowledged the platform’s inability to review the vast number of skill submissions, placing the onus on users to ensure the safety of the skills they deploy. This situation highlights the importance of user vigilance in the face of rapidly evolving threats.
Understanding the Risks
The rapid proliferation of these malicious skills underscores significant cybersecurity risks associated with open-source platforms. Users must be aware of the potential for misconfigured interfaces and the deep access that AI assistants like OpenClaw have to their systems. A multi-layered security approach is essential, including isolating the assistant in a virtual machine and restricting its permissions.
As this campaign continues to evolve, users should remain vigilant and proactive in monitoring their systems for unauthorized access or unusual activity. The use of tools like Koi Security’s free online scanner can help users assess the safety of skills before deployment, emphasizing the need for thorough checks in an increasingly complex threat landscape.
Key Takeaways
- Regularly check and update your OpenClaw skills to ensure they are from trusted sources.
- Use Koi Security’s free online scanner to evaluate the safety of skills before installation.
- Isolate the OpenClaw assistant in a virtual machine to limit its access to your system.
- Restrict permissions for OpenClaw to minimize potential damage from malicious skills.
- Monitor your accounts for any unauthorized access or unusual activity regularly.
Key Terms & Concepts
- OpenClaw: OpenClaw is an AI assistant that has undergone several name changes, including Moltbot and ClawdBot, and is designed to run locally.
- NovaStealer: NovaStealer is a type of malware that targets sensitive information such as cryptocurrency API keys and browser passwords.
- AuthTool: In this context, AuthTool is a malicious component disguised as a legitimate requirement for skills, used to deliver malware.
- ClawHub: ClawHub is the official registry for OpenClaw, where users can find and download skills for the assistant.
Read the original article here: https://www.bleepingcomputer.com/news/security/malicious-moltbot-skills-used-to-push-password-stealing-malware/
Your 5-Minute Securityish Brief
A weekly digest of cybersecurity news, phishing alerts, privacy tips, and emerging threats, simplified so anyone can understand what matters and why.
Securityish
Securityish explains cybersecurity, scams, data breaches, and privacy risks in simple language so you know what’s happening and how to protect yourself.
© 2026 Securityish
