recon-port-scan

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 0.90). The skill is high-risk: it provides comprehensive port-scanning tooling plus explicit IDS/IPS evasion, spoofing, decoy and stealth techniques (fragmentation, decoys, source-port/MAC/IP spoofing, idle/zombie scans, proxying, instructions to change IP if detected) that clearly enable covert, unauthorized reconnaissance and abuse, even though the bundled scripts do not contain obvious backdoor code, exfiltration endpoints, or credential-theft routines.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.80). The skill explicitly ingests and parses untrusted, third-party scan output (masscan JSON and nmap XML) — e.g., scripts/masscan_to_nmap.py, scripts/parse_nmap_xml.py, scripts/port_scan_storage.py and SKILL.md examples that pipe nmap/masscan output into these scripts — and uses those results to generate follow-up nmap commands and storage/reporting, so external content can materially influence subsequent tool actions.

MEDIUM W013: Attempt to modify system services in skill instructions.

• Attempt to modify system services in skill instructions detected (high risk: 0.70). The skill repeatedly instructs use of sudo (e.g., for stealth SYN scans, OS fingerprinting, and installing nmap) and includes evasion techniques that encourage privilege use/bypassing detection, so it pushes actions that can alter system state or require escalation even though it doesn't directly modify system files or create accounts.