sentry-setup-ai-monitoring

Monitor LLM calls and token consumption by configuring Sentry AI Agent Monitoring to automatically instrument your AI SDKs. The Skill detects installed libraries like OpenAI and Anthropic, then enables Sentry integrations to capture prompts, responses, and usage metrics end-to-end.

Sentry AI Agent Monitoring supports OpenAI, Anthropic, Vercel AI, LangChain, Google GenAI, and LangGraph. The Skill auto-detects which SDKs are installed in your JavaScript, TypeScript, or Python project and configures appropriate integrations for each.

Yes. Sentry AI Agent Monitoring works with existing LLM integrations by detecting your installed AI SDKs and enabling tracing without replacing your current setup. It adds observability for agent executions, tool usage, and token consumption alongside your existing code.

Set up automatic instrumentation by running the Skill, which detects your installed AI SDKs, verifies minimum versions, and presents detection results before applying per-SDK integration setups. It enables tracing and configures Sentry to capture LLM interactions automatically.

AI Agent Monitoring provides end-to-end visibility into agent workflows by tracking LLM calls, agent executions, tool usage, and token consumption. You see prompts, responses, and usage metrics in Sentry to understand how agents behave and consume resources.

Yes. The Skill provides guidance for manual instrumentation when automatic SDK detection isn't available. Manual setup lets you configure Sentry integrations directly in your code to monitor LLM interactions and agent behavior.