case-study-prep

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly fetches and ingests user-provided Google Docs via the google_drive_fetch step (Step 3) and then synthesizes that untrusted, user-generated content into answers, status decisions, and interview actions (Step 5), so third-party doc contents could materially influence agent behavior.