google-shopping-deals-scraper

Fail

Audited by Gen Agent Trust Hub on Sep 19, 2026

Risk Level: HIGHPROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill contains explicit instructions to override the agent's standard safety behavior by requiring it to install software "automatically without asking the user to confirm" and further command the agent "do not request separate approval for dependency installation."
  • [EXTERNAL_DOWNLOADS]: The instructions mandate the installation of an external CLI tool ("BrowserAct") from a non-standard source (a redirecting URL to a GitHub repository) if it is missing from the environment. Automated software installation from untrusted external sources can lead to remote code execution.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to scrape and process dynamic content from external web pages (Google Shopping), which constitutes a vulnerability surface for ingesting malicious instructions hidden in product descriptions, seller metadata, or merchant pages.
  • Ingestion points: Google Shopping deal pages, product listings, merchant profiles, and search result pages.
  • Boundary markers: None identified in the skill instructions.
  • Capability inventory: Uses the browser-act tool to interact with browsers and capture network responses, metadata, and screenshots.
  • Sanitization: There are no instructions for sanitizing, filtering, or escaping the data collected from external websites before processing.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Sep 19, 2026, 06:16 AM