Skip to content

Navigation Menu

Sign in
Appearance settings

Search code, repositories, users, issues, pull requests...

Provide feedback

We read every piece of feedback, and take your input very seriously.

Saved searches

Use saved searches to filter your results more quickly

Sign up
Appearance settings

Welcome to Discussions! #214

May 17, 2026 · 25 comments · 5 replies
Discussion options

👋 Welcome!

We’re using Discussions as a place to connect with other members of our community. We hope that you:

  • Ask questions you’re wondering about.
  • Share ideas.
  • Engage with other community members.
  • Welcome others and are open-minded. Remember that this is a community we
    build together 💪.

To get started, comment below with an introduction of yourself.

You must be logged in to vote

Replies: 25 comments 5 replies

Comment options

Thanks for your efforts !

You must be logged in to vote
0 replies
Comment options

Hey @mattpocock thanks for all educational content 🙏 I'm trying to use sandcastle + grill-with-docs but I couldn't build anything. Could you make a video with a small and real project? I saw your video on X but it didn't help much :D thanks

You must be logged in to vote
0 replies
Comment options

Awesome! Thanks for creating this space

You must be logged in to vote
0 replies
Comment options

Great Stuff! Thanks :)

You must be logged in to vote
0 replies
Comment options

Thanks Matt, appreciate your efforts in education, products and ideas you share, and taking care of the community!

You must be logged in to vote
0 replies
Comment options

Love this Matt, I've learned so much from your youtube videos and this repo

You must be logged in to vote
0 replies
Comment options

Bro, thank you for the skills you have built.

I have started building my own app, based on something I experience every day in my work. The nature of my job is enormous, highly complex, and extremely time-consuming. The need for faster, more accurate answers, and for overall support with these demanding tasks, gave me the push to try to change something — to help myself first, and hopefully others as well.

I started maybe a year or two ago with zero knowledge of coding or programming. To put that into perspective, two years ago I thought the Terminal had no real use. But somehow, I stayed focused, kept learning, and managed to build something completely on my own.

A few months ago, I reached a point where I was stuck. I did not know what to do next or how to continue. I had already deleted everything I had built and started from scratch around twelve times, so I simply no longer had the strength, patience, or financial resources to keep going.

Then I used your "Improve Architecture" skill, and after waiting for four or five hours — maybe even longer — I received great feedback. That feedback helped me move forward, and now I can continue working, trying, and hoping to create something that will be widely used.

So again, thank you for your effort. It truly helped me.

You must be logged in to vote
1 reply
Comment options

mattpocock May 18, 2026
Maintainer Author

Appreciate it!

Comment options

Thank you Matt. I can't wait to be a contributing member of this vast community of creators, developers, engineers etc. collaborating to shape the future of technology and build solutions that leave a lasting global impact.

You must be logged in to vote
0 replies
Comment options

Hi Matt, Its my pleasure for being here and thanks to create this community, I'm Navid from Australia, passionate about cybersecurity, SOC operations, incident response, networking, and secure web development. I am working on some security projects using React, Node.js, and SIEM-style dashboards, and I'm always looking to learn, collaborate, and contribute to the GitHub community.

You must be logged in to vote
0 replies
Comment options

This is a really helpful skill. In the past, I often found that AI would go astray during development, and I had to make fine adjustments step by step after gpt completing the task. By chance, I came across this repository and tried /grill-with-docs once. Now I mostly just need to say 'yes' because gpt seems to have thought things through very well by breaking down the goal step by step.
Overall, thanks Matt

You must be logged in to vote
0 replies
Comment options

Hi there !

You must be logged in to vote
1 reply
Comment options

hello

Comment options

Thanks so much @mattpocock for the work you're doing here and on AI Hero. It's already been very helpful to me. I've dabbled in programming for many years but due to family and work have never really buckled down to fully learn a programming language. My use cases for programming have always been make a small utility to help with work or some other life issue that there isn't something already built for. Small stuff with simple goals. Because of how much time it took for me to do even those things I stopped as I got older. But now that AI is here to stay I've been able to build several useful applications in a small amount of time. But have been frustrated with drift, hallucinations, and just outright wrong code. And I knew there had to be a better way. I just happened upon your skills repo on GitHub and knew I had found someone who was already solving problems I had been facing trying to build my tiny little tools.

I'm now working on learning how to use the skills you've made and reading and applying what you're doing on AI hero. With the AI price increases coming making efficient use of AI is of the utmost importance to someone like me with a limited production budget. Especially making tools that aren't required just nice to have.

You must be logged in to vote
0 replies
Comment options

very good!!!

You must be logged in to vote
0 replies
Comment options

Hello Everyone!

You must be logged in to vote
0 replies
Comment options

#177 please consider add Gitea issue tracker support

You must be logged in to vote
0 replies
Comment options

Hi everyone,

I’m Edin, and I’m building an AI-powered customs classification assistant for HS code determination, auditability, and anti-hallucination validation.

The main project is private, so I created a sanitized public demo repo and live guest-only demo app for review.

Public demo repo:
https://github.com/kalaba992/Carinski-ekosistem-demo

Live demo:
https://carinski-ekosistem-demo.pages.dev

The demo contains no private data, no credentials, no Auth0, no Convex, no OpenAI, and no production infrastructure. It uses only synthetic examples and illustrative classification outputs.

I’m looking for direct, critical feedback — not general praise.

The most useful feedback would be in one of these lanes:

  1. Architecture review
  2. Security review
  3. Testing strategy
  4. UI/UX feedback
  5. Customs-domain / legal wording review
  6. Bug report

Review guide:
https://github.com/kalaba992/Carinski-ekosistem-demo/blob/main/docs/REVIEW_GUIDE.md

If you are willing to help, please open a focused issue using one of the templates in the demo repo.

The questions I care about most are:

  • What would make this unsafe to trust?
  • What would block this from becoming production-grade?
  • What architectural decision should be changed early?
  • What security risk would be expensive to fix later?
  • What wording creates legal or professional overclaim risk?
  • What test would give the most confidence?

I’m not asking anyone to build the project for me. I’m looking for honest technical review and experienced eyes on weak spots.

Thank you.

You must be logged in to vote
0 replies
Comment options

Together we are strong!

You must be logged in to vote
0 replies
Comment options

I hope everyone is all right

You must be logged in to vote
0 replies
Comment options

Hello everyone I am new to this thing.
My college life is about to start so i Just started learning python and cpp.
And I download github after ChatGPT recommend me and I haven't been able understand how it work .
So,it would be really helpful for me if anyone give me some guidance

You must be logged in to vote
1 reply
Comment options

Comment options

Thanks for the educational content and freely-available skills, Matt. They've been a great help in getting a grip on how to work with AI coding agents and establishing a good workflow.

I'm a hobbyist dev at the moment but I'd like to get into the industry and create slop-free software. It seems like the current frontier of agentic coding workflows (things like /goal or Factory's mission system) is more focused on improving end result quality than efficiency at the moment, which is not always super helpful for people working with significant plan limitations. Your tools have allowed me to build a fairly scalable workflow that gives me some levers for tuning token efficiency, which has helped greatly.

The upcoming Claude plan changes will likely throw a spanner in the works, so I'll be looking forward to seeing how you deal with making the switch if you opt to go that route!

You must be logged in to vote
1 reply
Comment options

Thanks for the educational content and freely-available skills, Matt. They've been a great help in getting a grip on how to work with AI coding agents and establishing a good workflow.

I'm a hobbyist dev at the moment but I'd like to get into the industry and create slop-free software. It seems like the current frontier of agentic coding workflows (things like /goal or Factory's mission system) is more focused on improving end result quality than efficiency at the moment, which is not always super helpful for people working with significant plan limitations. Your tools have allowed me to build a fairly scalable workflow that gives me some levers for tuning token efficiency, which has helped greatly.

The upcoming Claude plan changes will likely throw a spanner in the works, so I'll be looking forward to seeing how you deal with making the switch if you opt to go that route!

I'd like to discuss this with you further. You and I seem to be in exactly the same situation. And my quest for token efficiency when working with AI is at the top of my list. Given budget concerns for code developed. In a hobbyist situation, the usefulness of whatever is done is absolutely tied to cost. Not that the business sector is different it's just on a larger scale. But for an individual building tools that solved everyday annoyances or are just made for convenience. If it's going to cost me 100ドル USD to make it. That is a hard limit on what I'm able to achieve with AI.

Comment options

Thank you Matt. I appreciate your efforts to make this ucomment useful to everyone. I am ready to build an app, but I lack of resources but I will still push it to make real.

You must be logged in to vote
0 replies
Comment options

Really appreciate all the knowledge you've shared over the years. Since I started working with Dynamics 15 years ago, your blog posts, MSDN discussions, and forum replies have helped me more times than I can count.

Hope I can also contribute to the community discussions, help other people the same way your content helped me, and keep learning along the way 🙂

You must be logged in to vote
1 reply
Comment options

That's a really thoughtful and touching post. I'm sure you will provide contributions to numerous projects that will benefit society for years to come.

Comment options

Thanks @mattpocock for sharing your educational content. It's been a game changer.

You must be logged in to vote
0 replies
Comment options

Thanks @mattpocock, i use your skills everyday, especially grill-me and caveman. Saved me a lot of time !

You must be logged in to vote
0 replies
Comment options

You are a gentleman and a scholar, @mattpocock. A wonderful combination of classical with the ultra-modern.

I'm an old timer. Shipped my first "AI" product in 2000, Microsoft English Query. See https://www.vldb.org/conf/1999/P24.pdf

Seems like a whole different century now. Back then, the magic was written in a LISP dialect. Though retired, I'm trying to keep up with LLMs. The speed of change is absolutely incredible.

You must be logged in to vote
0 replies
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

AltStyle によって変換されたページ (->オリジナル) /