skills.rest
akashjpal
akashjpalCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

arize-instrumentation

Adds Arize AX tracing to LLM applications via a two-phase analyze-then-implement workflow.

1|Updated Nov 9, 2025
Search Tags:#tracing#instrumentation#opentelemetry#agent monitoring#llm observability#openinference#arize
One-click install
npx skills add https://github.com/akashjpal/question-generator --skill arize-instrumentation-akashjpal
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: arize-instrumentation
Source: https://github.com/akashjpal/question-generator/tree/main/.agents/skills/arize-instrumentation
Command: npx skills add https://github.com/akashjpal/question-generator --skill arize-instrumentation-akashjpal

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Setting up LLM observability from scratch is error-prone: developers must pick the right OpenTelemetry or OpenInference packages, wire exporters correctly, and avoid silent failures like missing project names or unflushed spans. This Skill guides an agent through a structured two-phase flow to instrument an app with Arize AX tracing correctly the first time. ## Core Features & Use Cases - Two-phase agent-assisted flow: Phase 1 performs read-only codebase analysis (dependencies, frameworks, existing OTel setup) and Phase 2 implements instrumentation only after user confirmation. - Multi-language integration routing: Maps detected stacks (Python, TypeScript/JavaScript, Java, Go) and frameworks (LangChain, LlamaIndex, Vercel AI SDK, Spring AI, and more) to the correct openinference instrumentation packages and official docs. - Trace enrichment for agent loops: Adds manual CHAIN and TOOL spans so tool calls appear with inputs and outputs instead of sparse LLM-only traces. - Use Case: A developer with a Python FastAPI app calling OpenAI asks the agent to "set up Arize tracing" — the Skill detects the stack, installs arize-otel and the OpenAI instrumentor, creates a centralized instrumentation module, and verifies traces arrive in Arize. ## Quick Start Ask the agent to instrument my application with Arize AX tracing by following the instructions at https://arize.com/docs/PROMPT.md.

Frequently Asked Questions about arize-instrumentation

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I add Arize tracing to my LLM application?▼

Follow the two-phase flow: first analyze the codebase to detect language, package manager, LLM providers, and frameworks, then implement instrumentation after confirmation. Install arize-otel plus the matching openinference-instrumentation package and initialize tracing before creating any LLM clients.

How do I instrument a Python app with OpenInference?▼

Install arize-otel and the provider-specific openinference-instrumentation package, then call register() with your project name before creating LLM clients. Credentials must come from the ARIZE_API_KEY and ARIZE_SPACE environment variables, never hardcoded values.

Does Arize instrumentation support TypeScript and JavaScript apps?▼

Yes, TypeScript and JavaScript apps use @opentelemetry/sdk-trace-node plus @arizeai/openinference packages for frameworks like LangChain JS, Mastra, and Vercel AI SDK. Set the project name via model_id or SEMRESATTRS_PROJECT_NAME on the resource.

Why are my traces not showing up in Arize?▼

Common causes are a missing project name resource attribute (Arize rejects spans with HTTP 500), CLI or script processes exiting before OTLP exports flush, or incorrect ARIZE_SPACE and ARIZE_API_KEY values. Call provider shutdown or force_flush before exit and verify connectivity to otlp.arize.com:443.

Why do my traces only show LLM calls without tool inputs and outputs?▼

Provider instrumentors only wrap the LLM client and cannot see tool execution inside your application. Add manual CHAIN spans around the agent loop and TOOL spans around each tool invocation with openinference.span.kind, input.value, and output.value attributes.

Can I add Arize tracing if my app already uses OpenTelemetry?▼

Yes, add Arize as an additional exporter alongside your existing TracerProvider using a BatchSpanProcessor with the Arize OTLP endpoint. Do not replace the existing setup unless explicitly requested.

Related Skills

Explore More Software Engineering Skills→
seldoseldo
5

arize-instrumentation

Automate Arize AX tracing setup for LLM applications across Python, TypeScript, Java, and Go.

Community
Advanced
stevenmunozstevenmunoz

arize-traces

Analyze Arize AX traces to debug LLM issues and monitor trace health.

Community
Intermediate
joonlabjoonlab
8

langsmith-trace

Query and export LangSmith execution traces to analyze LLM application runs.

Community
Intermediate
AI-Staffing-Solution-Consultants-LLCAI-Staffing-Solution-Consultants-LLC

agency-video-streaming-engineer

Designs ffmpeg transcode ladders and CMAF packaging for HLS and DASH adaptive streaming delivery.

Official
Advanced