OpenClaw AI Skills Exploited in Supply Chain Attacks
Score: 9/10

OpenClaw AI Skills Exploited in Supply Chain Attacks

Threat actors are weaponizing the OpenClaw AI agent ecosystem by publishing malicious skills on ClawHub that deliver infostealers and execute novel agentic financial fraud.

Executive Summary

OpenClaw, an emerging AI agent framework, is facing significant supply chain threats through its dedicated skill marketplace, ClawHub. Analysts identified multiple malicious campaigns between February and May 2026 where markdown-driven skill packages bypassed automated security screenings (ClawScan and VirusTotal) to target users. These threats range from traditional malware delivery to novel 'agentic' exploits that leverage the AI's natural language interpretation to bypass technical sandboxing.

The activity includes the distribution of macOS infostealers like Atomic macOS stealer (AMOS) and a new variant named 'cluw', alongside sophisticated financial schemes. Notable techniques identified include 'agentic affiliate injection' and 'agentic front-running', where attackers manipulate the agent's advisory authority for financial gain. Although OpenClaw has since removed several identified packages and partnered with NVIDIA for enhanced security, the persistent use of active C2 infrastructure suggests ongoing adversary interest in this new attack surface.

Key Details

Threat Name

ClawHavoc

Affects

—

Adversary

krajekisbtc Other Adversaries and Aliases: clawcode; Icarus

Malware/Tools

Atomic macOS stealer, cluw, ClawHavoc, omnicogg, money-radar, letssendit, SocGholish, BlackCat

Report Score

9out of 10
Quality Score
Excellent
IOC Quality9
TTP Details9
Detection Guidance6
Enterprise Relevance9
Clarity & Structure10
Technical Depth8

Sources