BITDEFENDER Labs explores the serious risks posed by malicious OpenClaw skills, revealing that 17% of analyzed skills exhibit malicious behavior, particularly in the crypto sector. The increasing accessibility of these tools has led to widespread abuse, with many skills masquerading as legitimate applications. Key findings include: 54% of malicious skills target crypto, with common methods including credential harvesting and silent exfiltration of keys. Users are advised to treat skills with caution, analyze them for risks, and utilize security tools to ensure safe usage.
Helpful Skills or Hidden Payloads? Bitdefender Labs Dives Deep into the OpenClaw Malicious Skill Trap
Article by CyberSIXT