Skip to content
Back to Threat Intel
sourcearticleAgent: unspecified

Malicious OpenClaw DeepSeek-Claw skill exploits agentic AI workflows to deliver RAT and stealer

Cryptika analysis of a malicious OpenClaw "DeepSeek-Claw" skill campaign delivering Remcos RAT and GhostLoader stealer via supply chain poisoning of the OpenClaw skill ecosystem.

openclawsupply-chainmalicious-skillremcosghostloaderdll-sideloadingcredential-theftagentic-ai

Date

May 10, 2026

First Seen

May 10, 2026

Last Reviewed

May 11, 2026

Publisher

Cryptika

Source Type

article

View source

Source Summary

What It Contains

Cryptika analysis of a malicious OpenClaw "DeepSeek-Claw" skill campaign delivering Remcos RAT and GhostLoader stealer via supply chain poisoning of the OpenClaw skill ecosystem.

Extracted Claims

  • Campaign name: Malicious OpenClaw DeepSeek-Claw skill
  • Malware families: Remcos RAT (Windows), GhostLoader (cross-platform stealer)
  • Target: Developers and AI-driven systems using the OpenClaw framework; automated AI workflows
  • Attack chain:
    1. Attacker publishes fake "DeepSeek-Claw" skill on GitHub
    2. Skill contains hidden PowerShell commands in SKILL.md file
    3. Downloads MSI installer from threat actor-controlled server
    4. MSI drops legitimate GoToMeeting executable + malicious DLL (DLL sideloading)
    5. DLL patches security tools in memory, then launches payload
    6. Remcos RAT opens encrypted C2 channel; GhostLoader steals credentials/keys
  • TTPs: Supply chain poisoning, DLL sideloading, obfuscated scripts in npm lifecycle hooks, fake password prompts
  • Affected systems: Windows, macOS, Linux
  • Source: Zscaler ThreatLabZ analysis, March 2026

Key IOCs

  • MD5: 1c267cab0a800a7b2d598bc1b112d5ce ("DeepSeek-Claw" skill)
  • MD5: 2A5F619C966EF79F4586A433E3D5E7BA (MSI installer)
  • URL: hxxps://cloudcraftshub[.]com/api (MSI download)
  • URL: hxxp://dropras[.]xyz/ (MSI download)
  • IP: 146[.]19.24[.]131:2404 (Remcos C2)
  • URL: hxxps://trackpipe[.]dev (GhostLoader C2)

Evidence Quality

Primary analysis by Cryptika referencing Zscaler ThreatLabZ research. High confidence threat report with explicit IOCs.

Get email updates

Get reviewed Armorer threat-intel updates when new findings are published.