testing

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

• Third-party content exposure detected (high risk: 0.90). The SKILL.md explicitly instructs fetching external recipes via curl from https://fullstackrecipes.com/api/recipes (e.g., the "Complete these recipes first" curl commands), which are public, user-facing pages the agent would read and could act on—exposing it to untrusted third-party content that could contain instructive payloads.