AI is rapidly transforming how system administrators and developers interact with their environments. In 2025, we introduced the command line assistant, an optional tool available in Red Hat Enterprise Linux (RHEL) that allows users to ask questions and get natural language troubleshooting guidance right from the terminal.We're now announcing the next step in our AI-driven systems management journey—the availability of the goose AI agent in the RHEL extensions repository for RHEL 9.8 and RHEL 10.2. A project of the Agentic AI Foundation, goose is a flexible, open source AI agent that can be used with the Model Context Protocol (MCP) server for RHEL (developer preview) to unlock new use cases for AI-assisted administration.While the standard command line assistant provides a streamlined, read-only question and answer experience, gooseintroduces an extensible AI agent that helps enable context-aware system troubleshooting.2 client options for accessing the Red Hat hosted command line assistant backendWhen connecting to the Red Hat hosted command line assistant backend, administrators running RHEL now have 2 client pathways available. Depending on your requirements, you can choose between a traditional, streamlined Q&A tool or a newer, extensible tool with advanced features.1. The goose client (goose-redhat)For users looking to transition to a more powerful and capable assistant, we're introducing the goose-redhat package, available through the RHEL extensions repository on RHEL 9.8 and RHEL 10.2 systems. The goose-redhatpackage uses the open source goose AI agent to deliver advanced features such as support for MCP. The goose-redhat package configures goose to communicate with the same Red Hat provided command line assistant backend.2. The command line assistant client (command-line-assistant)The original command-line-assistant package is available within the standard RHEL 9 and RHEL 10 AppStream repositories. Invoked using the c command, this client is designed for a low-overhead, read-only Q&A experience. It lets users ask natural language questions, pipe command outputs for context, and attach log files to receive quick troubleshooting advice and command suggestions.The command-line-assistant package remains supported and available for use within the RHEL 9 and RHEL 10 systems. Please note, however, that we are not planning on adding new features to this client in the future. Selecting the right package for your environmentTo offer deployment flexibility, we're providing 2 distinct packages in the RHEL extensions repository.1. The goose-redhat package: integrated and optimizedIf you want an out-of-the-box agentic experience that is optimized for Red Hat environments, goose-redhat is your go-to.Installing this package provides streamlined access to Red Hat's command line assistant backend infrastructure, which can incorporate information from the RHEL documentation and Red Hat Knowledgebase articles. Additionally, goose-redhat has strengthened security settings so it will prompt for manual approval before it calls tools. 2. The goose package: Bring Your Own Model (BYOM)For users who would like to provide their own inference infrastructure, the standalone goose package is available. This allows you to connect goose to external large language model (LLM) providers or local inference servers. This package provides a similar experience to the upstream goose and includes the default upstream settings. How to get startedThe goose packages are distributed in the RHEL extensions repository, so you’ll first need to enable the repository on the RHEL system if you haven’t already. For RHEL 10:sudo subscription-manager repos --enable=rhel-10-for-$(arch)-extensions-rpmsFor RHEL 9:sudo subscription-manager repos --enable=rhel-9-for-$(arch)-extensions-rpmsTo install the goose-redhat package that is preconfigured with the Red Hat command line assistant backend:sudo dnf install goose-redhatOnce installed, simply open a new shell session, run goose, and ask a question.In this example, we ask if running RHEL 6-based containers on a RHEL 9 host is supported. Goose uses the Red Hat command line assistant backend to answer the question, and provides a link to the source document on the customer portal.$ goose
Supercharge RHEL troubleshooting with agentic AI: Introducing goose
Learn about the goose AI agent, now available in the Red Hat Enterprise Linux (RHEL) extensions repository for RHEL 9.8 and RHEL 10.2. This open source AI agent can be used with the Model Context Protocol (MCP) server for RHEL to unlock new use cases for AI-assisted administration, providing context-aware system troubleshooting. Get started with the goose-redhat package today.









