share

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

• Third-party content exposure detected (medium risk: 0.30). Outsider-authored free text can enter via the user-supplied “report”/file in SKILL.md step 1, and the required runtime workflow then ingests that report by generating a slides.json whose fields are rendered into slides.html/PNGs by scripts/render_carousel.py (which reads slides.json at line 148 and embeds all slide text into the output at lines 90-124 and 139).