instrument-llm-observability

What problem does it solve?

This Skill addresses the challenge of adding tracing and observability to LLM or AI-agent applications, ensuring every run is inspectable with details like prompts, tool calls, token usage, latency, and cost per step.

Core Features & Use Cases

• Tracing Integration: Offers vendor-neutral OpenTelemetry for tracing LLM applications.
• Cost and Performance Monitoring: Captures cost per span and latency data.
• Custom Instrumentation: Allows manual spans for custom tool/retrieval steps.
• Use Case: When debugging an LLM app in production or implementing cost and performance tracking, this Skill can be triggered to add the necessary tracing and observability features.

Quick Start

Run the 'instrument-llm-observability' skill to add production-grade tracing to your LLM application.