1. Habit
  2. Skills
  3. Arize Dataset skill
arize-ai · 3K installs

Arize Dataset skill
Runs by itself, new test cases added.

A growing, verified evaluation dataset.

Install the Arize Dataset skill and make it a habit: follow https://gethabit.dev/skills/arize-ai/arize-dataset.md

  1. 1Copy
  2. 2Paste into Claude Code, Codex or Cursor
  3. 3It runs by itself

3K installs65 starsMIT licenceSource on GitHub ↗

Arize Dataset

by arize-ai · 3K installs

The skill

The arize dataset skill creates, manages and queries Arize datasets with the ax CLI: create from files, append examples, export versions and verify completeness.

Only when you ask.

The habit

A growing, verified evaluation dataset.

↻ Runs by itselfnew test cases added

How the habit works.
You don't have to.

Three questionsWhen does it start? How does it check its work? What does it remember?

01

It starts by itself

new test cases added. Runs in GitHub Actions or Claude Code routine.

  • ↻ new test cases added
  • new examples file committed

02

It checks its work

export count matches server. If a check fails, it tells you.

  • ✓new examples have the same field names as existing ones
  • ✓ax datasets append exits with success
  • ✓exported row count matches the server example_count

03

It remembers

dataset version notes. The next run reads it first: the latest dataset version and its example count.

loops/arize-eval-dataset-loop/run-log.md

One run, step by step.
Then it stops.

Stop ruleStops after one append and one verified export per run, with at most one retry.

  1. 01

    Find dataset

    Your agent looks up the dataset with ax datasets get by name or ID.

  2. 02

    Check schema

    It exports one example and compares its field names with the new data to catch typos.

  3. 03

    Append

    It appends the new examples with --json or --file, leaving out server fields like id.

  4. 04

    Export all

    It exports the dataset with --all.

  5. 05

    Verify count

    It compares the exported row count with the server's example_count and re-exports if they differ.

  6. 06

    Hand off

    It points to the arize-experiment skill to run evaluations on the updated dataset.

What it may do.
Nothing more.

Safe by defaultIt writes only in its own folder and asks you before anything goes out.

Is it for you?
Use it when it fits.

Quick checkRead both lists. If the left one sounds like you, the Arize Dataset skill helps.

Use it when

  • ✓ creating a golden dataset
  • ✓ appending test examples
  • ✓ exporting a dataset version

Skip it when

  • – you do not use Arize
  • – you only need to run experiments

Questions

Arize Dataset skill, answered.

Why check field names before appending?

A misspelled field silently creates a new column, so the skill compares keys first.

Why export with --all?

An export without --all stops after one page, so the count check would fail.

How do I install the Arize Dataset skill?

Copy the prompt at the top into your agent. It installs the skill from arize-ai/arize-skills and sets up the habit. Or run npx skills add arize-ai/arize-skills --skill arize-dataset.

Is the Arize Dataset skill free?

Yes. It is published under MIT. Habit is free too.

Arize Dataset. Now as a habit.

Install the Arize Dataset skill and make it a habit: follow https://gethabit.dev/skills/arize-ai/arize-dataset.md

  • Free
  • No account
  • Undo in one line

All arize-ai skills →More code skills →

Works in

  • Claude Code
  • Codex
  • Cursor
  • ChatGPT
  • Claude app

Made by arize-ai, published under MIT. Habit links to the original and never changes it. Is this your skill?

Copying didn't work here. Select the line and copy it by hand.