Skills
skills/cookjohn/ieee-skills/ieee-navigate-pages/Gen Agent Trust Hub

ieee-navigate-pages

Pass

Audited by Gen Agent Trust Hub on Mar 15, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it ingests untrusted data from the IEEE Xplore website to determine the current state. * Ingestion points: SKILL.md extracts DOM content (result counts and active page buttons) via browser scripts. * Boundary markers: None. * Capability inventory: navigate_page and evaluate_script (which can execute arbitrary browser-side code). * Sanitization: None.
  • [COMMAND_EXECUTION]: The skill performs dynamic code execution within the browser using evaluate_script. This includes an initScript designed to evade bot detection by redefining navigator.webdriver.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 15, 2026, 07:11 AM