infrastructure-enumeration

Warn

Audited by Gen Agent Trust Hub on Jul 6, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill invokes numerous CLI tools (e.g., nmap, snmpwalk, dnsrecon) and specifically requires elevated privileges via sudo mount to interact with NFS exports.
  • [DATA_EXFILTRATION]: Includes commands to attempt retrieval of sensitive files like /etc/passwd via TFTP and instructions to store discovered credentials and vulnerability evidence in local files.
  • [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection by processing untrusted data from network services. (1) Ingestion points: Tool outputs from network probes (DNS, LDAP, SNMP) described in SKILL.md. (2) Boundary markers: None present; external data is not delimited or marked as untrusted. (3) Capability inventory: Subprocess execution of network tools across all sections and local file writes to the engagement/ directory. (4) Sanitization: No evidence of data sanitization or validation of tool outputs before processing.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Jul 6, 2026, 09:15 AM
Security Audit — agent-trust-hub — infrastructure-enumeration