amazon-kindle-wishlist

Pass

Audited by Gen Agent Trust Hub on Jul 12, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill utilizes browser automation tools (Chrome DevTools or Computer Use) to perform actions on Amazon.co.jp. The instructions are confined to wishlist management without external network calls or data exfiltration.
  • [SAFE]: Explicit instructions prevent the agent from performing purchase-related actions such as 'Add to Cart' or 'Buy Now', mitigating financial risk.
  • [SAFE]: The skill implements a verification process for items, requiring matches for titles and ASINs before any removal action is taken.
  • [SAFE]: Analysis of the data ingestion surface (Amazon pages) indicates that while the agent processes external content, the instructions provide sufficient boundary markers and sanitization steps (ASIN/title validation) to mitigate indirect prompt injection risks.
  • [SAFE]: No obfuscation, unauthorized privilege escalation, or persistence mechanisms were detected in the instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 12, 2026, 06:19 AM
Security Audit — agent-trust-hub — amazon-kindle-wishlist