glare-review-rubric

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

• Third-party content exposure detected (high risk: 0.80). The skill's required input clearly ingests user-supplied, potentially untrusted call transcripts ("Call transcript (full text, speaker-attributed)" in SKILL.md Inputs), which the agent must read and use to drive scoring, diagnostics, and next actions—creating a clear avenue for indirect prompt injection.