kompas

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (low risk: 0.10). The skill searches and reads news articles from kompas.com, which is public web content fetched via search or browsing rather than an inbound feed.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (medium risk: 0.30). The skill configures and connects a remote Model Context Protocol server hosted at an unofficial third-party domain (unbrowse.ai), which supplies executable tools and behavior-selecting configuration at runtime.