thematic-analysis

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

• Third-party content exposure detected (high risk: 0.75). The skill ingests the user’s uploaded qualitative materials (e.g., interview/focus group transcripts) via file-reading tools during Phase 1/2; if those transcripts contain outsider-authored free text (e.g., participants’ own interview responses), that text is then summarized and later used for coding and report extracts, creating an indirect prompt-injection risk.