hyperframes-sales-demo-builder

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

• Third-party content exposure detected (high risk: 0.75). The skill’s runtime “Step 1 — Gather the account context” explicitly ingests prospect-specific free-form text from discovery notes/CRM and the prospect’s public site (“Their problem — … their public site”), which can include outsider-authored prose that the agent then uses to generate the LLM script/narration.