web-content-fetcher

Pass

Audited by Gen Agent Trust Hub on Apr 1, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill references external utility services, specifically r.jina.ai (Jina Reader) and defuddle.md, to facilitate web content extraction and Markdown conversion. These are well-known services commonly used for providing LLM-friendly content from dynamic or advertisement-heavy websites.
  • [PROMPT_INJECTION]: The skill facilitates the ingestion of content from arbitrary third-party URLs, which introduces a surface for indirect prompt injection.
  • Ingestion points: Untrusted data enters the agent context via content fetched from user-specified or searched URLs in SKILL.md.
  • Boundary markers: The instructions do not explicitly mandate the use of delimiters or instructions to ignore embedded commands within the fetched content.
  • Capability inventory: The agent processes the fetched content for display or transformation tasks, using tools like WebFetch and WebSearch.
  • Sanitization: There is no explicit requirement for the agent to sanitize or filter the content before processing.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 1, 2026, 09:01 AM