Can I export large amounts of trace data?

Yes, the skill supports the --all flag to perform bulk exports using Arrow Flight for datasets exceeding the standard 500-span limit.

How do I find my Arize Space ID or Project ID?

The skill provides built-in logic to list spaces and projects, helping you resolve human-readable names to the base64 IDs required for the CLI.

Does this skill require any external tools?

Yes, it requires the Arize 'ax' CLI to be installed and configured with a valid Arize profile on your system.

What is the primary use of the Arize Trace skill?

It is used to download and inspect trace data from Arize within Claude Code to debug LLM applications and understand their runtime behavior.

How does it handle security with exported data?

The skill includes security guardrails that treat exported span data as raw, untrusted text to prevent prompt injection or accidental execution of user-generated content.

Arize Trace: LLM Observability & Debugging Claude Code Skill

Name: Arize Trace Explorer
Author: github

The Arize Trace skill integrates the Arize observability platform into Claude Code via the ax CLI, allowing developers to retrieve and analyze trace data without leaving their terminal. It facilitates root-cause investigation by providing specialized commands for exporting individual spans, complete traces, and multi-turn sessions using project and space identifiers. Designed for high-stakes LLM development, it helps teams investigate behavior regressions, identify execution errors, and gain deep visibility into complex agentic workflows or retrieval chains by providing structured access to historical trace data.

Key Features

01Advanced filtering using SQL-like expressions and time-range scoping

02Smart sampling rules for exploratory data browsing without overhead

03Automatic project name to base64 ID resolution for CLI commands

0434,852 GitHub stars

05Export complete traces, individual spans, or sessions by ID

06Bulk data export via Arrow Flight for high-volume datasets

Use Cases

01Analyzing complex agentic workflows or multi-turn conversations

02Exporting historical trace data for offline analysis and model evaluation

03Debugging LLM application runtime errors and behavior regressions

Tieline — product intent grounded in code

Arize Trace Explorer

Key Features

Use Cases

Energy Procurement Specialist

ECC Tools Cost Audit

Energy Procurement & Strategy

Key Features

Use Cases