q-educator

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

• Third-party content exposure detected (high risk: 0.80). The skill's required demo and assignment workflows (see SKILL.md Demo Outline and references/demo_example.md) explicitly instruct selecting and classifying social-media/project comments (e.g., "Select 20 comments from the dataset", "operationalize the construct 'skill-based discourse' in social media comments"), which are untrusted, user-generated third‑party content that the agent is expected to read and that can materially influence analysis choices and subsequent tool use.