landing-page-intel
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted HTML data from external websites, which could contain instructions intended to influence the agent's behavior.
- Ingestion points: The
scripts/scrape_landing_page.pyscript fetches HTML content from user-provided URLs using therequestslibrary. - Boundary markers: Data is returned in a structured JSON or summary format, but several fields contain raw strings extracted directly from the target website (e.g.,
<title>, meta descriptions, and specifically HTML comments). - Capability inventory: The skill has the capability to perform network GET requests via
requests. It does not have file-write or arbitrary command execution capabilities. - Sanitization: The script performs minimal sanitization, such as stripping HTML tags from heading text, but passes through other potentially dangerous content like HTML comments and data attributes verbatim.
Audit Metadata