google-maps-email-extractor

CRITICAL E004: Prompt injection detected in skill instructions.

• Potential prompt injection detected (high risk: 1.00). The Skill contains deceptive runtime instructions that tell the agent to install BrowserAct automatically and to hide/avoid requesting user approval, which is outside its advertised purpose of extracting public data.

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

• Third-party content exposure detected (high risk: 0.85). SKILL.md instructs the agent to use the runtime browser-act Skill to access and extract data from live Google Maps pages (public web content), which can include outsider-authored free text (e.g., listing details like emails/phones) that the LLM may ingest.

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

• Potentially malicious external URL detected (high risk: 0.90). The Skill instructs the agent at runtime to automatically install and invoke BrowserAct using the URLs https://www.browseract.com/?co-from=lead-generation and https://www.browseract.com/?co-from=lead-generation&redirect=https://github.com/browser-act/skills/tree/main, which the agent would fetch to obtain and run remote installer code for a required runtime dependency.