manus

Warn

Audited by Socket on Mar 29, 2026

1 alert found:

Security
SecurityMEDIUM
SKILL.md

SUSPICIOUS. The skill’s capabilities broadly fit its purpose, but trust and data-flow integrity are weakened by a personal-repo publisher and an API endpoint mismatch with official Manus documentation. It delegates prompts, files, and potentially sensitive work to a powerful third-party autonomous agent, so overall security risk is medium-high even without clear malware behavior.

Confidence: 87%Severity: 72%
Audit Metadata
Analyzed At
Mar 29, 2026, 07:38 AM
Package URL
pkg:socket/skills-sh/mvanhorn%2Fclawdbot-skill-manus%2Fmanus%2F@0933267565455bcfbd436abf7661d1a247345c52