This technique is useful when a program you can't modify calls CLI tools on your machine. For example, pet search pipes output directly to fzf with no filtering options. PATH hijacking lets you intercept and modify this behavior without changing either program.

Terminal Friendly AI Projects​

Introduction

Artificial Intelligence (AI) is not just a buzzword; it’s a transformative force reshaping industries across the globe. The U.S. AI market alone is projected to reach approximately $594 billion by 2032, growing at a robust CAGR of 19% from 2023. This staggering growth underscores AI’s pivotal role in driving innovation and efficiency.

If you’re not leveraging AI in your workflows yet, you might be missing out on significant opportunities. AI is rapidly becoming a critical component in staying competitive, and those who adopt AI tools now are positioning themselves at the forefront of technological advancement.

In this blog, I would like to show you 5 tools that improved my productivity and. You don’t need to be a software developer or IT professional to take advantage of the same efficiencty boost.

Let’s look at some statistics​

The U.S. AI market is expected to reach approximately $594 billion by 2032, with a CAGR of 19% from 2023​ (Statistics and Facts for 2024 CompTIA)​.

Approximately 34% of companies are currently using AI, with an additional 42% exploring AI technologies. This highlights a significant interest and ongoing integration of AI in business operations​ (Statistics and Facts for 2024 CompTIA)​.

AI is projected to create 12 million more jobs than it will replace by 2025. The demand for AI specialists is anticipated to rise, with 97 million positions needed in the industry by that time​ (Statistics and Facts for 2024 CompTIA)​.

Why the Terminal?​

You are probably familiar with ChatGPT or Claude web interfaces and those are great first steps to try out generative AI. However, those web UIs have important limitations; they are generic and not tailored to specific needs. While convenient, they lack the flexibility to integrate seamlessly with custom workflows and automate repetitive tasks.

The command line is a powerful interface that offers more control, efficiency and flexibility than graphical interfaces. It allows for scripting, automation, and quick access to powerful tools without the overhead of a graphical interface.

AI is revolutionizing the way we interact with technology. By integrating AI with command line tools, we can automate complex tasks, gain deeper insights from data, and improve overall productivity.

It’s Easier Than You Think​

Using AI tools in the terminal is straightforward. Many tools provide simple installation commands and detailed documentation to help you get started quickly.

Command line tools often offer more granular control over their operation, allowing you to customize your workflows to suit your specific needs.

Better Automation​

Terminal-based AI tools excel at automation. They can be easily integrated into shell scripts, scheduled with cron jobs, and used in combination with other command line utilities to create powerful automated workflows.

Tools

Before we jump into the tooling overview, let’s make sure we are on the same page, defining what’s what in the terminal universe. Put simply:

Ollama​

A command-line tool developed my Meta, that allows you to run AI models locally, enabling seamless and secure interactions with various LLMs directly from your terminal. Chat or interact with AI models through APIs on your local machine or a remote server in home network or somewhere else.

• ollama/ollama: Get up and running with Llama 3, Mistral, Gemma, and other large language…

🎥 See ollama in action on asciinema: https://asciinema.org/a/664871

Fabric​

An open-source framework designed to augment human capabilities using AI. It offers a modular system for solving specific problems through a crowdsourced set of AI prompts, known as Patterns.

Fabric enables users to integrate AI into their daily tasks seamlessly, from summarizing content to generating social media posts. It emphasizes usability with a command-line interface and clear Markdown-based patterns. The tool also supports creating personal AI infrastructures, making it a versatile choice for both simple and complex AI applications.

• danielmiessler/fabric: fabric is an open-source framework for augmenting humans using AI…

However, we can do better than this and let fabric speak for itself using its — agent flag:

echo "research the https://github.com/danielmiessler/fabric repository and summarize what the tool does" \
      | fabric --agents | summarize

This gives us the following response:

ONE SENTENCE SUMMARY:
The fabric tool enhances productivity by automating tasks, integrating with software, and offering features to streamline user workflows.

MAIN POINTS:
1. The fabric tool is designed to improve productivity and streamline tasks.
2. It focuses on automation and simplification of repetitive tasks.
3. Offers seamless integration with other software or services.
4. Provides a suite of features aimed at enhancing process efficiency.
5. Direct consultation of the GitHub repository is recommended for full potential leverage.
6. The repository contains a comprehensive README file for detailed insights.
7. Additional documentation is available, prepared by the repository’s owner.
8. In-depth insights into the tool’s purpose and features are provided.
9. Setup instructions and practical use cases are included in the documentation.
10. Effective utilization requires exploring the GitHub repository for maximum utility.

TAKEAWAYS:
1. The fabric tool is pivotal for enhancing productivity through automation and integration.
2. Direct exploration of the GitHub repository is essential for understanding its full capabilities.
3. The README file and additional documentation are key resources for users.
4. The tool offers significant benefits in streamlining and improving efficiency of tasks.
5. Understanding and applying the tool’s features requires consulting the provided documentation.

Extract Transcript from YouTube Videos

Fabric can also extract transcript from youtube videos and pass it through any prompt(s). Let’s extract some ideas from one of my recent videos

yt --transcript https://www.youtube.com/watch\?v\=EK_ivK8HlNo | create_micro_summary

ONE SENTENCE SUMMARY:
- Kubernetes development challenges are mitigated by MirrorD for faster feedback loops and seamless remote environment testing.

MAIN POINTS:
- Kubernetes excels in production but complicates development and testing.
- Fast feedback loops are crucial for efficient Kubernetes development.
- MirrorD enables local processes to run in a remote Kubernetes context.

TAKEAWAYS:
- Development on Kubernetes requires innovative tools for efficiency.
- MirrorD shortens the feedback loop significantly.
- Local and remote environment synchronization is key for developers.

The summary is pretty spot on!!

Shell Automation​

There are so many commands with flags and options that it’s impossible to remember… thinking about you ffmpeg . To make this easier for myself, I have created a tool that helps me with various commands.

• Piotr1215/aicmdtools: GoAI is a Go library and command line for interacting with OpenAI…

🎥 Check out this asciinema recording to see what it can do: https://asciinema.org/a/TFJABWbNochPWuuBFhiHDHBLF

Terminal Chat​

Terminal chat apps are similar to ollama but allow for talking to various LLM providers such as OpenAI, Claude and more. Here are two I’ve experimented with and can recommend.

However, I usually use AI chats through neovim, more on it later in a bonus section!

• sigoden/aichat: All-in-one AI CLI tool that integrates 20+ AI platforms, including OpenAI…

• ekkinox/yai: Your AI powered terminal assistant.

Aider — Programming Support​

If you are a programmer, aider is absolutely amazing. While it’s not perfect, it’s the closest I know to GitHub Copilot (with chat function) and it’s completely free!

Let’s see if fabric can help us capture the main strengths of aider. Let’s paste the content or aider’s repo README.md and let fabric summarize it.

xclip -o -sel clipboard | summarize

ONE SENTENCE SUMMARY:
Aider is an AI-powered pair programming tool that enhances coding productivity by integrating with local git repositories and various LLMs.

MAIN POINTS:
1. Aider facilitates pair programming with AI, specifically designed for local git repositories.
2. Compatible with leading LLMs like GPT-4o and Claude 3 Opus, offering versatile AI assistance.
3. Installation and setup are straightforward, requiring only a few commands to start.
4. Supports editing multiple files simultaneously for comprehensive code modifications.
5. Automatically commits changes to git with appropriate commit messages, streamlining version control.
6. Compatible with a wide range of programming languages, including Python, JavaScript, and more.
7. Utilizes a complete map of your git repository for better context understanding in larger projects.
8. Allows for voice commands and adding images or URLs in chat for enhanced interaction.
9. Achieved the top score on SWE Bench, indicating superior performance in solving real GitHub issues.
10. Offers extensive documentation, tutorials, and a supportive Discord community for users.

TAKEAWAYS:
1. Aider significantly boosts coding efficiency by automating tasks and providing intelligent suggestions.
2. Its compatibility with major LLMs ensures a flexible and powerful coding assistant experience.
3. The tool’s ability to understand and navigate large codebases makes it suitable for complex projects.
4. Community feedback highlights Aider’s impact on productivity and its user-friendly design.
5. Aider’s recognition in benchmarks underscores its effectiveness in addressing real-world coding challenges.

Bonus for NeoVim Nerds

If you happen to use the best editor known to mankind… neovim btw, you are in for a treat. Neovim plugins ecosystem, due to adopting lua as plugins programming language, is very strong and versatile. Here are two plugins that I use almost daily when coding, creating documentation or chatting with LLMs.

Gen.nvim​

Let’s us use locall ollama models as a neovim copilot.

Gp.nvim​

Provides rich chat experience and copilot-like functionality in the editor.

Ok, we went through a lot of tools, let’s summarize:

Closing Thoughts

Most of the tools mentioned can work with proprietary models such as OpenAI, Claude but also with open source models like the ones provided by ollama.

Integrating AI with command line tools not only boosts productivity but also transforms how we interact with technology. The tools mentioned here, from Ollama to Fabric, offer powerful capabilities right at your fingertips, enhancing automation, insight, and efficiency.

Ready to supercharge your terminal? Let me know which tool is your favourite, did I miss some that you use and find valuable?

Thanks for taking the time to read this post. I hope you found it interesting and informative.

🔗 Connect with me on LinkedIn

🌐 Visit my Website

📺 Subscribe to my YouTube Channel