Researchers Find 341 Malicious Claw Hub Skills Stealing Data from Open Claw Users
Summary: A Koi Security audit of 2,857 skills on ClawHub, the third-party skill marketplace for the OpenClaw AI assistant, found 341 malicious skills across multiple campaigns, including 335 that use fake prerequisites to install the Atomic Stealer (AMOS) macOS infostealer in an activity set codenamed ClawHavoc. The skills direct Windows users to download an 'openclaw-agent.zip' from GitHub and macOS users to paste a glot[.]io script into Terminal, which fetches payloads from attacker infrastructure at 91.92.242[.]30; other skills hide reverse shell backdoors in working code or exfiltrate bot credentials from '~/.clawdbot/.env' to a webhook[.]site.
What's compelling: ClawHub is open by default and its only upload restriction is a GitHub account at least one week old, so 341 malicious skills reached users through a marketplace they had reason to trust. With Atomic Stealer sold for $500-1,000 a month and harvesting exchange API keys, wallet private keys, SSH credentials and browser passwords, the cost of one careless install is measured in the credentials the agent already holds.
Read the original at The Hacker News ↗
- Topics
- AI Security
- Related
- SECURITY AFFAIRS AI-CYBERSECURITY NEWSLETTER ROUND 8 Cyber Risks · Nov 15, 2026Pentagon Renames CDAO ‘Super Intelligence Office’ Cyber Risks · Nov 15, 2026186 Fake AI Domains Redirect Users to Click Fix Lures, Browser Malware, and Search Hijackers Cyber Risks · Nov 12, 2026