speech-build

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

• Third-party content exposure detected (high risk: 0.70). The skill's references/source_code.md explicitly instructs using "web_fetch" to retrieve raw code from public URLs (e.g., https://raw.githubusercontent.com/googleapis/...), which requires the agent to fetch and read untrusted third-party content from the open web for deep inspection, exposing it to indirect prompt injection.