mantis-meta-agent
Warn
Audited by Socket on Sep 14, 2026
1 alert found:
SecuritySecuritySKILL.md
MEDIUMSecurityMEDIUM
SKILL.md
SUSPICIOUS. The skill is internally aligned with a security-review orchestrator, and there are no external installer, credential-harvesting, or third-party proxy signals. However, it grants broad autonomous control over a long-running review-and-patch pipeline, includes destructive local file operations, and coordinates offensive security tooling and many subagents, making the overall operational risk high despite no clear evidence of malware.
Confidence: 90%Severity: 74%
Audit Metadata