dev-server-port-detection

Pass

Audited by Gen Agent Trust Hub on Mar 18, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill includes a bash script scripts/detect_port.sh that executes standard system utilities such as lsof, netstat, and curl to identify listening ports on localhost. This is a standard procedure for development environments.
  • [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface by instructing the agent to parse terminal output from build tools which may include untrusted content. (1) Ingestion points: Terminal logs described in references/port-detection-patterns.md. (2) Boundary markers: Absent. (3) Capability inventory: Shell script execution in scripts/detect_port.sh. (4) Sanitization: Absent for terminal output parsing, though file-based input is filtered for numeric values.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 18, 2026, 02:41 AM
Security Audit — agent-trust-hub — dev-server-port-detection