amara

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

• Third-party content exposure detected (high risk: 0.80). The skill instructs the agent to use Membrane actions (e.g., membrane action run / get-subtitles, get-video, list-video-urls) to fetch Amara content—often user-generated subtitles and video metadata—from a third-party platform and to read/use those results as part of its workflow, which could allow indirect prompt injection.