qa-only

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

• Third-party content exposure detected (high risk: 1.00). The SKILL.md explicitly instructs the agent to visit arbitrary target URLs and browse the app under test (e.g., Phase 3 "Orient" uses $B goto and Phase 4 "Explore" uses $B goto and follow site links), meaning it fetches and interprets third‑party web content as part of its workflow.