skills/wgpsec/aboutsecurity/iox-proxy/Gen Agent Trust Hub

iox-proxy

Warn

Audited by Gen Agent Trust Hub on Sep 23, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill provides pre-constructed shell commands for the iox utility, including specific arguments for port forwarding, SOCKS5 proxy creation, and network scanning via nmap and proxychains.
  • [COMMAND_EXECUTION]: The skill includes instructions to modify the system configuration file /etc/proxychains.conf by appending proxy settings using shell redirection (>>).
  • [DATA_EXFILTRATION]: The core functionality described involves setting up reverse proxies and tunnels specifically designed to route internal network traffic to external systems, which facilitates bypassing network security controls and exfiltrating traffic.
  • [INDIRECT_PROMPT_INJECTION]: The skill establishes a vulnerability surface by creating a command-execution pipeline for external network operations.
  • Ingestion points: Instructions in SKILL.md guide the agent to interpret user requests for network pivoting.
  • Boundary markers: Absent. The skill does not include delimiters or warnings to ignore potentially malicious embedded instructions in the network traffic it facilitates.
  • Capability inventory: Includes execution of the iox binary, system file modification (echo to /etc/proxychains.conf), and network scanning (nmap).
  • Sanitization: Absent. There is no validation or filtering of the parameters passed to the networking tools.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Sep 23, 2026, 04:24 AM