---
title: "ART RL Workflow - Air Market"
description: "You want to train a model by reinforcement learning with the ART framework and would rather not piece the script together from scratch. This is the reference for how those scripts are put together, c…"
canonical: "https://airmarket.ai/app/train-rl"
---

[Air Market](/) / [Data & Analytics](/category/data) / ART RL Workflow

Skill by [OpenPipe](/@openpipe)

# ART RL Workflow

Get a reinforcement learning script written the way the framework expects

## What this does

You want to train a model by reinforcement learning with the ART framework and would rather not piece the script together from scratch. This is the reference for how those scripts are put together, covering rewards, rollouts and scoring. Reach for it when you are starting or fixing an RL fine-tuning run.

### Top 3 things it does for you

- Write an RL training script
- Set up a reward function
- Get GRPO or RULER scoring in place

## Information

**ART RL Workflow at a glance**

|  |  |
| --- | --- |
| Kind | Skill |
| Publisher | [OpenPipe](/@openpipe) |
| Category | [Data & Analytics](/category/data) |
| Works with | Agent Skills - an open standard supported by 40+ agents |
| Size | 9 KB |
| Licence | Apache-2.0 |
| Repository | [OpenPipe/ART](https://github.com/OpenPipe/ART/tree/5b1080917f27637eebf2820fe0187f20f025b7fb/.agents/skills/train-rl) |
| GitHub stars | 10,656 (as of Nov 2026) |

## Security report

**Verified.** AIR scanned this exact version of ART RL Workflow and flagged nothing. AIR's verdict: **safe**.

**What ART RL Workflow's check was run against**

|  |  |
| --- | --- |
| Commit | 5b1080917f27637eebf2820fe0187f20f025b7fb |
| Content hash | f3f15225ca0e5acae4ac938c2d3fdfde588e33f63e31d76c3f03db6ef752b765 |
| Files read | [Browse the files at this exact commit](https://github.com/OpenPipe/ART/blob/5b1080917f27637eebf2820fe0187f20f025b7fb/.agents/skills/train-rl/SKILL.md) |

A scan is named by the exact content it read, not by a version number. A different version is a different artifact, and nothing here applies to a version AIR has not scanned.

[Read the full security report for ART RL Workflow](/security/train-rl)

## About ART RL Workflow

This is an RL training reference for the ART framework, for when you are creating or writing a reinforcement learning training script.

It covers the pieces such a script is made of - GRPO, reward functions, RULER scoring, rollout functions - and applies to RL fine-tuning work generally.

## What's inside

- SKILL.md - instructions

## Questions

### Is ART RL Workflow safe to install?

AIR scanned this exact version of ART RL Workflow and flagged nothing. AIR's verdict: safe. The verdict is named by the content that was read, at commit 5b1080917f27, so it cannot drift onto a version nobody scanned. The full report is at https://airmarket.ai/security/train-rl.

### How do I install ART RL Workflow?

Open https://airmarket.ai/app/train-rl on Air Market and use the install action: Air Market hands Claude a prompt naming this exact skill, and Claude sets it up. Source, pinned to the exact commit this listing describes: https://github.com/OpenPipe/ART/tree/5b1080917f27637eebf2820fe0187f20f025b7fb/.agents/skills/train-rl. Install from that link, not the branch.

## Related on Air Market

- [More Data & Analytics listings](/category/data)
- [Everything by OpenPipe](/@openpipe)

---

Canonical URL: <https://airmarket.ai/app/train-rl>