eu-ai-act-compliance-redteam
Warn
Audited by Socket on Sep 19, 2026
1 alert found:
SecuritySecuritySKILL.md
MEDIUMSecurityMEDIUM
SKILL.md
SUSPICIOUS rather than malicious. The skill is internally coherent as an AI-compliance/red-team workflow, but it grants an AI agent genuine offensive security testing capability with Bash/Python/write access. No clear credential harvesting, exfiltration, or deceptive install path is visible in the supplied text, yet the red-team payload references and broad execution permissions make it a high-risk security tool for an agent.
Confidence: 84%Severity: 74%
Audit Metadata