dns-network
Fail
Audited by Gen Agent Trust Hub on Apr 14, 2026
Risk Level: HIGHCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides instructions to execute commands with root privileges using
sudo, includingsudo tcpdumpfor packet capture andsudo iptablesfor firewall management. - [COMMAND_EXECUTION]: The skill directs the agent to modify the
/etc/hostssystem configuration file usingsudo teeandsudo sed. Such modifications can redirect system-wide network traffic or be used for persistence. - [PROMPT_INJECTION]: The skill presents an attack surface for indirect prompt injection by interpolating user-supplied hostnames and domains into shell commands without sanitization. Evidence Chain: 1. Ingestion points: User-provided hostnames, IP addresses, and domain lists (
domains.txt) inSKILL.md. 2. Boundary markers: Absent. 3. Capability inventory: Execution ofdig,curl,nc,tcpdump, andsudoacross the skill. 4. Sanitization: No input validation or escaping instructions are provided for handling untrusted data.
Recommendations
- AI detected serious security threats
Audit Metadata