sglang

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

• Third-party content exposure detected (high risk: 0.85). The required runtime workflow is an OpenAI-compatible chat/completions server that ingests the caller-provided messages/user_query text into the prompt context; that caller is an outsider relative to the operating user, so their free-form text can become LLM-readable context (indirect prompt injection risk via user input).