Shadowfetch: the Debian Linux distribution focused on local AI

Shadowfetch: An Innovative Linux Distribution Integrating Local AI

Shadowfetch is a new Linux distribution based on Debian, distinguished by its integration of local artificial intelligence (AI). It aims to simplify the installation and use of language models (LLMs) and AI agents while remaining true to the open-source spirit of Linux. However, it is crucial to note that Shadowfetch is still in the early stages of its development.

The idea of combining AI with Linux elicits varied reactions. For some purists, it may seem like heresy, but others see it as an inevitable evolution. Linus Torvalds, the creator of Linux, even suggested that those who oppose AI should consider distancing themselves from it. In this context, a developer has set out to transform Debian into a platform dedicated to local AI, potentially useful for discussions about AI, application development, or other tasks where AI can add value.

The Challenges of Installing Shadowfetch

The Shadowfetch distribution is still in its initial development phase, apparently led by a single developer, which could slow progress. Despite this, it already shows promising potential. However, I do not yet recommend Shadowfetch for any use beyond curious experimentation.

During installation, I encountered numerous obstacles. The installation consistently failed, forcing me to run Shadowfetch from a live instance in RAM, which is not optimal for an AI workstation. With each attempt, the installation would hang midway, and after reconnecting, the entire process would reset, even though the disk partition had been created.

A Distribution with Multiple Options

Despite these difficulties, using Shadowfetch in RAM allowed me to glimpse the developer's vision. This distribution aims to be extremely flexible, allowing users to choose from various configuration profiles before installation. These profiles include options like Creative Studio, Developer Workstation, Gaming, Office and Communication, Local AI Lab, and Maker and Hardware. You can combine several of these profiles to create the workstation that perfectly fits your needs.

During configuration, it is also possible to customize the AI according to your preferences. You can opt for a private local AI, choose your language model, and select between two agents: Hermes and OpenClaw. This ability to customize the AI is one of the most interesting aspects of Shadowfetch, although it is not yet installable on a hard drive.

Integrated Tools and Applications

Once configured, Shadowfetch offers a handy AI tool, the Shadowfetch Assistant, which allows interaction with the chosen agents or models. Shadowfetch uses Ollama in the background, making it particularly efficient.

I was pleasantly surprised by how quickly Shadowfetch's AI tool was able to create a Linux Qt application. The AI chat tool operates without slowing down the system, generating responses rapidly. I even challenged the AI to design an application for selecting a daily outfit, and it responded faster than I had anticipated. Compared to Ollama on my main desktop, Shadowfetch proved to be significantly faster.

Preinstalled Applications Based on Selected Profile

The preinstalled applications vary depending on the selected profile. For example, the Office and Communication profile includes tools like Calibre, LibreOffice, PDF Arranger, Thunderbird, and Syncthing Web UI. By choosing the Creative Studio profile, you will have access to applications such as GIMP, Audacity, Ardour, Easy Effects, Kdenlive, and Shotcut. The diversity of available applications is impressive and can meet many needs.

Who is Shadowfetch For?

Shadowfetch may appeal to users looking for a Linux distribution with a dedicated AI studio, facilitating the installation of agents like Hermes and OpenClaw. Although AI features are not yet integrated into applications like GIMP and LibreOffice, this could be considered in the future.

This distribution is ideal for those who wish to leverage AI without resorting to cloud services, staying on a local system. I hope the developer can resolve the installation issues, allowing for a deeper exploration of Shadowfetch. For now, I recommend using it only in live mode to avoid unnecessary frustrations.