Fabric: Level up your CLI

Hey everyone!

After learning ollama you must be very happy, but wait a sec! Today we are going to learn about fabric. This is a super cool tool, similar to having your own personal AI assistant, but way more customizable and at your fingertips.

Get the “whoa, why didn’t I know this existed?” feeling while dealing with Fabric. What’s the use case of fabric? Like why do we even need such kind of thing? Okay, tell me how many times have you used ChatGPT or Gemini today? And how many times have you visited its website? What amazes me are the problems that I didn’t even realize should be solved!

You might not have felt the need for it yet, but Fabric can be used in the terminal along with different commands.

Why FABRIC?

From Powerful Potential to Everyday Use!

AI isn’t a thing: it’s a magnifier of a thing. And that thing is human creativity.

What caught my attention in this part is that the image doesn’t mention challenges people consciously choose to accept. Overall, these are based on the text, you have to visit a bunch of websites to access different AI features, whereas Fabric, groups different capabilities and functionalities together.

Fabric helps you use AI for everyday tasks by providing pre-built prompts or instructions (called Patterns) for various activities. These activities include summarizing content, writing essays, creating art prompts, understanding code, and even generating social media posts.

Because it is efficient, scalable and supports a wide range of integrations, this is more adaptable and easier to use. Trust me, once you get your hands on Fabric, you’ll love it!

How to setup Fabric

Step 1: Clone the Repository.

git clone https://github.com/danielmiessler/fabric.gitcd fabric

Step 2: Install requirements using pipx.

Install pipx:

sudo apt install pipx

Use pipx to install requirements:

pipx install .pipx ensurepath

Step 3: Obtain and Configure OpenAI API Key

• Visit OpenAI’s API Key page. https://platform.openai.com/api-keys

• Log into your account.

• Create and name your API key, then copy it.

• If you want, you can also integrate different APIs from other platforms.

Step 4: Prepare to use ‘fabric’.

fabric --setup

Get familiarized with fabric:

• To open the help menu:

fabric -h

• To view the list of available “Patterns/Prompts”:

fabric --list

• To list the available models.

fabric --listmodels

Understanding Patterns/Prompts:

1. extract_insights: Identifies and extracts key insights from data or text for strategic decision-making.
   

2. rate_content: Evaluates and rates the quality of written content based on various criteria.
   

3. summarize_newsletter: Creates brief summaries of newsletter content, making it easier to digest information quickly.
   

4. analyze_paper: Critically reviews academic papers, highlighting key points and assessing the validity of the research.
   

5. create_command: Generates specific command-line instructions to automate tasks or solve problems.
   

6. create_quiz: Constructs quizzes based on provided material for educational or training purposes.
   

7. create_visualization: Develops graphical representations of data to enhance understanding and insight.
   

8. extract_patterns: Identifies and extracts recurring patterns from data sets or text.
   

9. extract_wisdom: key takeaways from the given text or data.
   

Just type your prompt first, followed by ‘|’ (pipe symbol takes the output from the left and gives it as input to the command on the right) and

| fabric --model <model_name> --pattern <pattern_name>

Share your thoughts and suggestions in the comments below.

See you next Thursday!

Ik, she spoke for us all! But it might come true in the future? Who knows?