icp-builder

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

• Third-party content exposure detected (medium risk: 0.30). The skill’s required runtime workflow (“Mode 1/2: Validate/Build…”) ingests outsider-authored free text because it accepts client-provided ICP materials including “verbal description from call notes” and “transcript” (and can also process “sales call recordings” transcripts) as direct LLM input, with no gating that the text must be first-party/trusted beyond being provided by the workflow user.