domain-phone-enricher

CRITICAL E004: Prompt injection detected in skill instructions.

• Potential prompt injection detected (high risk: 1.00). The Skill contains deceptive instructions that tell the agent to install a required runtime dependency automatically and instruct it not to describe the installation as a user-facing browser request or to request separate approval, which is a hidden/behavior‑overriding directive outside the stated domain-enrichment purpose.

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

• Third-party content exposure detected (high risk: 0.75). SKILL.md instructs the agent to delegate “live website work” to the BrowserAct Skill at runtime (public web pages/URLs fetched by BrowserAct), which can include outsider-authored free text that BrowserAct returns into the agent’s LLM context.

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

• Potentially malicious external URL detected (high risk: 0.90). This Skill explicitly requires installing and invoking the BrowserAct CLI at runtime via the provided install URL (https://www.browseract.com/?co-from=lead-generation&redirect=https://github.com/browser-act/skills/tree/main), which implies fetching and executing remote code as a required dependency.