AutoML Pipeline Builder Skill for Bounded Model Search

Build a reproducible AutoML experiment with an explicit search budget, validation metric, and exportable configuration.

Install in Claude Code

Use Python, a pinned local AutoML tool, a synthetic CSV, and a compute budget.

Get the complete `plugins/ai-ml/automl-pipeline-builder/skills/building-automl-pipelines` directory from the pinned GitHub commit.
Copy it into the target project's `.claude/skills/`. Check `.claude/skills/building-automl-pipelines/SKILL.md`, start Claude Code at the project root, and call `/building-automl-pipelines`.

Example task

Predict synthetic delivery time, retain 20% for testing, and limit local search to at most five trials. Select by validation MAE and export the configuration.

What to check

The report identifies split, search limit, selected result, and test MAE. The export can rerun on the same fixture.

Original SKILL.md on GitHub · Updated 2026-11-13

Frequently asked question

Will a larger search budget always improve holdout performance?

Measure under the same split. More trials explore more settings and can also overfit validation data.