---
title: "Onboard your repository to Testing Farm with Claude Code"
description: "Configuring automated testing for Red Hat Enterprise Linux and Fedora projects on Testing Farm usually requires writing boilerplate CI workflows and TMT metadata by hand. The testing-farm-onboarding AI agent skill for Claude Code automates this by generating .fmf test metadata and CI pipeline definitions for GitHub Actions, GitLab CI, and Red Hat internal GitLab, covering both public and enterprise environments. It interactively prompts for API tokens, target operating system composes, and trigger events such as pull requests or pushes, and supports matrix builds across multiple distributions. The skill analyzes existing test suites like tox, creates valid TMT plans, and integrates with tmt plan lint and tmt test lint for validation before committing. Installation is done by cloning the sclorg/ai-agents repository and symlinking the skill into the project’s .claude/skills directory, reducing onboarding from hours to minutes."
---

# Onboard your repository to Testing Farm with Claude Code

[Red Hat](https://yomu.fyi/company/red-hat) · Petr Hracek · Nov 2, 2026

**Type:** Tutorial

## Summary

Configuring automated testing for Red Hat Enterprise Linux and Fedora projects on Testing Farm usually requires writing boilerplate CI workflows and TMT metadata by hand. The testing-farm-onboarding AI agent skill for Claude Code automates this by generating .fmf test metadata and CI pipeline definitions for GitHub Actions, GitLab CI, and Red Hat internal GitLab, covering both public and enterprise environments. It interactively prompts for API tokens, target operating system composes, and trigger events such as pull requests or pushes, and supports matrix builds across multiple distributions. The skill analyzes existing test suites like tox, creates valid TMT plans, and integrates with tmt plan lint and tmt test lint for validation before committing. Installation is done by cloning the sclorg/ai-agents repository and symlinking the skill into the project’s .claude/skills directory, reducing onboarding from hours to minutes.

## Context

Setting up Testing Farm for a repository requires writing TMT metadata and CI pipeline definitions manually, which involves friction with API tokens, metadata formats, and platform-specific workflows.

## Approach / What changed

The testing-farm-onboarding AI agent skill for Claude Code automates the generation of TMT metadata (.fmf files) and CI pipeline definitions. It is installed by cloning the sclorg/ai-agents repository and symlinking the skill into the project’s .claude/skills directory. The skill is invoked via a slash command or direct prompt, then interactively guides the user through selecting the Git platform (GitHub, GitLab, or Red Hat internal GitLab), choosing trigger events, specifying target operating system composes, and providing API tokens. It analyzes existing test suites and generates valid TMT plans and workflow YAML files that can be validated with tmt lint commands.

## Takeaways

- The skill automates TMT metadata and CI pipeline generation for repositories on GitHub, GitLab, and Red Hat internal GitLab, supporting both public and enterprise environments.
- It interactively prompts for required configuration such as API tokens, target OS composes, and trigger events, and supports matrix builds across multiple distributions.
- Generated TMT plans and test definitions can be validated using tmt plan lint and tmt test lint before committing, ensuring schema correctness and script path validity.

## FAQ

### What platforms does the Testing Farm onboarding skill support?

It supports repositories hosted on GitHub (via GitHub Actions), GitLab (via GitLab CI), and Red Hat internal GitLab (gitlab.cee.redhat.com).

### How do I install the skill?

Clone the sclorg/ai-agents repository and symlink or copy the testing-farm skill into your project's .claude/skills directory or user-level skills directory.

### How can I validate the generated TMT metadata?

Use \`tmt plan lint\` to validate test plans and \`tmt test lint\` to validate individual test definitions. The linter checks schema, required keys, and script path existence.

**Tags:** [Automation](https://yomu.fyi/topic/automation), [CI/CD](https://yomu.fyi/topic/ci-cd), [Claude](https://yomu.fyi/topic/claude), [Red Hat](https://yomu.fyi/topic/red-hat), [Testing Farm](https://yomu.fyi/topic/testing-farm)

- Source: [Red Hat](https://developers.redhat.com/articles/2026/11/02/onboard-your-repository-testing-farm-claude-code)
- Source URL: https://developers.redhat.com/articles/2026/11/02/onboard-your-repository-testing-farm-claude-code
- Ingested by Yomu: 2026-11-02T15:00:17.497Z

[Read original post](https://developers.redhat.com/articles/2026/11/02/onboard-your-repository-testing-farm-claude-code)
