Unlock AI with Fabric Patterns

It’s a great tool. I built a web interface for it & built an aggregator website utilising it as well. I use it to summarise blogs/channels and generate links on specific topics so I can see if it’s worth reading the original article/video Main catch with the YT is the query limit with the API

Matt, here’s a suggestion- see if it is possible to have a live chat with the author, that would really be interesting. Great video as always, will try out fabric today! 👍

I’m excited to learn more about Fabric!

I started writing my own Python tools to do this, including downloading and concat'ing multiple transcripts for upload to chatbots for discussion. So cool to stumble across your channel!

Yes please create more Fabric videos with use cases students, creators and data analysts can make use of. Great video thank you! Subscribed

This was hugely beneficial! Thank you so much!

I'm glad you've gotten into fabric, I use it very often and have found it to be great.

I started using it recently. I use to to output markdown to my Obsidian vault. I created special version of save that also calls fabric to create the front-matter tags for it so my vault can auto connect related content

Using it for a while. In love with it 😅 Have created new patterns for it too.

I would like to access fabric from a CMD or powershell under windows. To run fabric you have to use wsl under windows. It would be great if you could address this topic in one of your upcoming videos!

another banger! Thank you

Thanks. You did it again.

It's funny; I was examining Fabric with a friend on Tuesday, and my takeaway was, "The tool itself may not be the greatest, but the value truly lies in these prompts."

First youtuber I've seen recommned fish, I love it. I'd just like if more documentation took it into account as setting env variable etc it a bit different.

Hi Matt, one nice thing to do at the end when you take you favorite drink, expose a scrolled window at the side to show all the code you've written during that video. Sounds good?

YES! It's all about the prompts. The confusion over fabric, comes from how they bury the lead with the framework and tools they've built, which is cool but I'm not a YouTuber, I don't work in Cyber security, I'm not a Mac user, I'm not a Christian, and I don't play DnD. I just need to know how to write good prompts.

I’ve created a prompt library tailored to my workflow since I’m a civil engineer and landscape architect, which are niche fields. Many AI tools are designed for general use, not specifically for those of us who build and maintain our infrastructure—true “civil” engineering, from the word “civilization,” the mother of all engineering and landscape architecture, henceforth. While there are many interesting tools out there, they often don’t provide practical help for my work. I’m also developing niche software to better support my profession since no one in the tech field understands anything more than virtual reality. There are too many outsiders to the profession that miss the mark when developing software so I guess I am going to have to step up to the plate.

i am really interested in how you are using fabric patterns with your code. are the scripts already released in that repo ?

@technovangelist Thank you for the video. I have Ollama installed on Windows 11 system. Got Fabric installed, put API for Google/YT in setup. Cool to get the video summary, but when I pipe to "extract wisdom" I get unrecognized command.

I'm baffled at the praise Fabric is getting tbh; They seem to have spent more time on buzzwords like 'fabric,' 'loom' and 'weave' than they have actual functionality. The directory separator in the "save" script is a hard-coded forward slash... At the end of the day it seems like little more than a very limited-use prompts library, especially compared to something like Langchain. Perhaps I'm missing something, but after trying it a few times, I just cannot find a use-case that isn't better served by more established, more flexible, and more professionally-written tools. The "extract wisdom" and youtube use-case seems to be getting a good majority of the attention. Langchain's summary chain, rag-driven chat, and a prompt for extracting ideas work together to provide a much better solution for that use-case.