Arize Dataset

Handles dataset operations for Arize's evaluation and experimentation workflow. You can create datasets from CSV, JSON, JSONL, or Parquet files, append new examples, export everything back out, and manage versions. The CLI is `ax datasets`, and it works across spaces with name or ID lookups. One nice touch: you can pipe data directly via stdin instead of saving temp files, and there's an auto-escalation rule that catches truncated exports at 500 rows and reruns with the unlimited flag. Useful when you're building test sets, managing golden datasets, or need to version your eval data alongside experiments.

skills.sh installs: 3.2K
GitHub stars: 72
last updated: Aug 21, 2026

View on GitHub

Install

$ npx -y skills add arize-ai/arize-skills --skill arize-dataset --agent claude-code

Installs into .claude/skills of the current project.

Files

Arize Dataset Skill

SPACE — --space flags accept a space name (e.g., my-workspace) or a base64 space ID (e.g., U3BhY2U6...). Find yours with ax spaces list.

Concepts

Dataset = a versioned collection of examples used for evaluation and experimentation
Dataset Version = a snapshot of a dataset at a point in time; updates can be in-place or create a new version
Example = a single record in a dataset with arbitrary user-defined fields (e.g., question, answer, context)
Space = an organizational container; datasets belong to a space

System-managed fields on examples (id, created_at, updated_at) are auto-generated by the server -- never include them in create or append payloads.