competitor-analysis

Warn

Audited by Gen Agent Trust Hub on Nov 22, 2026

Risk Level: MEDIUMMETADATA_POISONINGEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [METADATA_POISONING]: The skill metadata identifies the author as 'aaron-he-zhu', which contradicts the author information provided by the platform ('filipas123'). This discrepancy could be misleading regarding the origin and safety of the skill.\n- [EXTERNAL_DOWNLOADS]: The skill suggests running an 'npx' command to install additional components from an unverified third-party repository ('aaron-he-zhu/seo-geo-claude-skills'). This pattern facilitates the execution of unverified remote code on the user's system.\n- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from external competitor websites, which is a common vector for indirect prompt injection.\n
  • Ingestion points: Competitor URLs and their website content are fetched and processed to analyze SEO strategies.\n
  • Boundary markers: The instructions lack specific delimiters or warnings to treat external content as untrusted data.\n
  • Capability inventory: The skill uses SEO tools and generates reports, providing a path for injected instructions to influence agent output.\n
  • Sanitization: There are no requirements for sanitizing or validating the content retrieved from external URLs.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Nov 22, 2026, 01:11 AM