red-run-legacy

SUSPICIOUS/HIGH-RISK but not confirmed malware. The skill is internally aligned with its stated pentest-orchestrator purpose, yet that purpose is itself an offensive security orchestration role with broad delegated execution, dynamic skill loading, target interaction, credential handling, and post-exploitation workflows. No clear credential-harvesting or deceptive exfiltration endpoint is shown, but the AI-agent attack surface and real-world offensive capability make the overall security risk high.