sentry-setup-ai-monitoring

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

• Third-party content exposure detected (high risk: 0.80). This skill explicitly enables capturing user- and model-generated content from AI SDK calls (e.g., recordInputs/recordOutputs, include_prompts, and gen_ai.request.messages set from LLM messages), so it ingests untrusted third-party content coming from AI providers and user inputs.