traffic-analysis
Warn
Audited by Snyk on Feb 16, 2026
Risk Level: MEDIUM
Full Analysis
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (high risk: 0.90). The skill instructs use of packet-capture and analysis commands (e.g., tcpdump -A 'tcp port 80', ngrep 'GET' port 80, and tshark -Y "http.request" -T fields ...) that display and extract arbitrary HTTP/DNS network traffic from the open web, so the agent would read untrusted third-party content.
Audit Metadata