sentry-setup-ai-monitoring

The artifact is a benign, purposeful instrumentation guide to enable Sentry AI/LLM observability. It does not contain obfuscated or malicious code nor supply-chain install-execute patterns. The main security concern is privacy and accidental data exfiltration: enabling prompt/output capture or using high trace sampling rates without governance can send PII and secrets to Sentry. Operators should treat opt-in capture settings carefully, audit dependencies for unexpected AI SDKs (to avoid inadvertent auto-instrumentation), and ensure proper data handling policies, retention limits, and access controls on the Sentry project.