Arize Instrumentation

This is a two-phase instrumentation assistant that reads your codebase first, then adds Arize AX tracing after you confirm the plan. It detects your stack (Python, TypeScript, Java, Go) by scanning dependencies and imports, figures out which LLM providers and frameworks you're using, then routes to the right OpenInference auto-instrumentation packages or manual OpenTelemetry spans. The analysis phase is read-only and won't touch your code until you approve. Useful when you want observability on an existing LLM app without hunting through integration docs yourself. It prioritizes framework-native tracing over provider-specific when available and never hardcodes credentials into generated code.

skills.sh installs: 1.1K
GitHub stars: 41K
last updated: Aug 21, 2026

View on GitHub

Install

$ npx -y skills add github/awesome-copilot --skill arize-instrumentation --agent claude-code

Installs into .claude/skills of the current project.

Files

Arize Instrumentation Skill

Use this skill when the user wants to add Arize AX tracing to their application. Follow the two-phase, agent-assisted flow from the Agent-Assisted Tracing Setup and the Arize AX Tracing — Agent Setup Prompt.