gws-gmail

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

• Third-party content exposure detected (high risk: 0.90). The SKILL.md explicitly includes commands that fetch and read user emails (e.g., +read, +triage, +watch which "read a message and extract its body or headers" and "stream them as NDJSON"), so the agent will ingest untrusted, user-generated third-party email content from Gmail that could contain instructions influencing subsequent actions.