15-social-listening-global

Pass

Audited by Gen Agent Trust Hub on Jun 19, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [SAFE]: The skill provides a strategic and operational framework for brand monitoring and crisis management. It does not include any executable code, scripts, or automated configurations.
  • [NO_CODE]: The entire file consists of Markdown-formatted text, including tables, workflows, and templates, without any associated scripts (Python, Node.js, Shell).
  • [PROMPT_INJECTION]: The skill is designed to ingest untrusted data from social media and review platforms (Category 8: Indirect Prompt Injection surface).
  • Ingestion points: Social media mentions, Reddit discussions, TikTok comments, and Google Reviews defined in SKILL.md.
  • Boundary markers: No specific delimiters or "ignore instructions" markers are provided for the data processing steps.
  • Capability inventory: None. The skill does not define any subprocess calls, file-writes, or network operations in script form.
  • Sanitization: No sanitization or validation methods are specified for the external content.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 19, 2026, 03:33 PM
Security Audit — agent-trust-hub — 15-social-listening-global