competitive-ads-extractor

Pass

Audited by Gen Agent Trust Hub on Jul 16, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill consists entirely of markdown instructions and examples, containing no scripts, binaries, or automated installation commands. It relies on the agent's pre-existing tools for its operations.\n- [COMMAND_EXECUTION]: The skill provides instructions for the agent to save analysis reports and screenshots to local directories (e.g., ~/competitor-ads/). These operations are transparent, limited to the user's home directory, and necessary for the skill's functionality.\n- [DATA_EXFILTRATION]: Network activity is described as accessing public advertising libraries (Facebook, LinkedIn). No indicators of sensitive data harvesting or exfiltration to unauthorized third-party servers were found.\n- [INDIRECT_PROMPT_INJECTION]: As the skill involves the analysis of external, user-controllable ad text, a potential surface for indirect prompt injection exists. However, this is inherent to any tool that processes web content and is not an active exploitation in this skill.\n
  • Ingestion points: Scraped ad headlines, body copy, and metadata from external platforms.\n
  • Boundary markers: None provided in the instruction set.\n
  • Capability inventory: Local file system write access for report generation.\n
  • Sanitization: Not specified; the agent is expected to apply its own content filtering and safety protocols.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 16, 2026, 04:33 PM
Security Audit — agent-trust-hub — competitive-ads-extractor