test-case-gen

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

• Third-party content exposure detected (high risk: 1.00). The skill explicitly instructs the agent to navigate to and capture rendered DOM from user-provided URLs (see rules/ref-input-sources.md and the "Generate from URL" example in SKILL.md), meaning it fetches and parses arbitrary third‑party web pages whose content is then used to drive test-case generation and decisions.