skills/cookjohn/sd-skills/sd-download/Gen Agent Trust Hub

sd-download

Pass

Audited by Gen Agent Trust Hub on Mar 5, 2026

Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it processes user-provided identifiers (PII) and URLs to navigate and interact with external websites.
  • Ingestion points: The user-supplied PII and URL parameters are directly interpolated into browser navigation commands and script templates.
  • Boundary markers: No explicit boundary markers or instructions to ignore embedded commands are present in the prompt interpolation logic.
  • Capability inventory: The skill utilizes powerful browser automation capabilities including navigation (navigate_page), arbitrary JavaScript execution (evaluate_script), and element interaction (click).
  • Sanitization: There is no evidence of input validation or sanitization for the PII or URL strings before they are used to build executable browser scripts.
  • [EXTERNAL_DOWNLOADS]: The skill facilitates the download of PDF documents from ScienceDirect, which is a well-known academic publisher. The download operation is consistent with the skill's stated purpose.
  • [COMMAND_EXECUTION]: The skill executes JavaScript within the browser context to automate the extraction of PDF links and the handling of security verification widgets. This includes script-based evasion of bot detection by modifying the 'navigator.webdriver' property.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 5, 2026, 06:45 AM