Preliminary research

ClawMutiny: We Audited 1,620 OpenClaw Skills

AI-collected research leads through 2 October 2026, including bounded month-by-month reviews of selected social and community sources from January through September. Unranked, incomplete, not community-vetted, and subject to change.

Measures instruction-layer threats across 1,620 OpenClaw skills and compares the results with a conventional malware-oriented scanner. Manual review of disagreements highlights prompt-delivered credential theft, remote instruction updates and anti-scanner text that clean install-time analysis misses.

Record

Published by

Oathe

In the archive

• Collected in the 2026 preliminary research sweep

Tags

• ai-agent
• prompt-injection
• supply-chain
• measurement-study
• scanner-bypass
• credential-theft
• owasp-a03-2021
• owasp-a06-2021

This page is the archive's own catalogue record. The research is the work of Oathe, first published at the original source. Preserved copies are kept so the citation survives its host.