google-gemini-embeddings

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

• Third-party content exposure detected (high risk: 0.85). The skill’s runtime RAG flow can ingest outsider-authored free text from the /ingest endpoint (documents[].text) and later feeds it into the LLM as context in generateResponse() (text: \... Context:\n${context}\n\nQuestion: ${query}\n\nAnswer:``), which is indirect prompt-injection risk.