cast

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

• Third-party content exposure detected (medium risk: 0.65). Runtime path: the skill ingests user-provided content (URL via WebFetch / pasted text / file Read) and then embeds that content as readable HTML/text into {{CONTENT_HTML}} / {{BODY_HTML}} for rendering, so any outsider-authored free text included in that input can flow into the LLM context via the agent’s analysis step (Step 1/Step 2) before HTML generation.