Garratt Campton 🍄

If you want a link, PM.

AI - Hot Take: Stop calling repos "Awesome Blah Blah" when it's mostly crap. Calling something "Awesome Agent *" When it's literally just a library of f@*king 30 line prompts is utter garbage. He's me making complex RAG systems with entire Documentation API's, making actual working "AGENTS" that don't make mistakes. Like how the f@*k are these guys getting 20k stars for this garbage? At least give me a 1mb greppable file to go with your poor prompting. I think we are doomed as a species. 😵‍💫

That didn't even include Niche Agents/skills I've built, crypto, gaming, local services.

I removed Gteam and made it private. 5 stars, sitting for a month. Lawyers, Accountants, local/external/on-page SEO's, Social-media, Systems, Ideas-Man, 28 agents. but over 250 files of knowledge. 330mb.

@techn0eacc @thsottiaux Yeah this, Anthropic made it clear with their leaked files they make it so Claude doesn't give helpful suggestions. I noticed this months ago between the two and would preference using Chat -> Codex. Make it better than the competition, it already is.. but make it better :)

@thsottiaux I wish Codex would step in with questions when the user wasn't specific enough about their request. Like what'd be the purpose of the app user wants to create.

What are we obviously not getting right with Codex?

@thsottiaux Customisable shortcuts below the prompt window.

@EMostaque @thsottiaux Emad, Surely you're not still using ChatGPT? I literally stopped at this thread to say "Shouldn't you be asking us what we are obviously not getting right with ChatGPT?"

@thsottiaux It is difficult to see/know how xhigh (with agents?) compares to gpt pro standard / extended so power users likely waste time/tokens pasting between ChatGPT and codex Guidance or benchmarks or pro skill would sort this

For decades people coded on Windows as the most popular OS (this is not up for debate). I'm thinking now with so many switching to macs for Ai use, we are also seeing a significant shift in app building that targets Mac first as OS choice. Generally speaking. What do you think?

@dotta @AntoineLescun nice nice, the UXDesigner as a chief role urks me. I keep trying to make, CFO, CTO, COO, CMO, and CPO. and the CEO will still hire the UX as a chief anyway. instead of just having UX under CPO.

@AntoineLescun Totally agree. The next versions will allow multiple teams, any roots, and have a parallel human org structure

Paperclip has faced some skepticism for being org-chart based Won't the infinite, all-knowing agents create their own organizational structures? Isn't the org chart anachronistic? Let me give you "7 Reasons why human org Charts Help Agents Run Things (CHART)" Starting with..

@WeedPorns Wat... AI is racist AF.

This is what AI was built for

Me: Oh yeah lets do this build in public thing... Goes to my 20 year old fully verified YouTube account:

I will start doing proper walkthrough installs and recording my day to day . Its fun to see a site come alive so quickly.

Full setup of Agent-Pixels site took 1hr 37 mins. From "npx install mantlekit" to "done" Wire up Supabase, Stripe, Brevo, Google, Github OAuth. Stripping dummy content and writing blog posts + home and about page Not sure if that's faster than WordPress? agent-pixels.com

@sudoingX I've got a 4070 and basically 0 usable ram, what model should I go for?

if you are running local ai or thinking to start, if i could give you one single piece of advice it is this: choose your agentic harness carefully. it matters more than the model. i have lost count of how many people have dm'd me saying their local model is "dumb" or "broken" or "not as good as the cloud one." then they switch from openclaw or some other bloated framework to hermes agent and the same model suddenly works. just clean tool calls and the agent doing the thing it was supposed to do. hermes agent is the best general purpose agent i have used in 2026. drives my single 3090 with qwen 3.6 27b dense q4, drives my dgx spark with nemotron omni q8, and the same harness handles coding, research, video editing, automation, anything you point it at. packed with skills out of the box (browser tools, code, github, jupyter, multimodal, more than i have used yet), full tool calling that holds across long sessions, persistent memory, sub agents. if you tried local ai once or twice and gave up because it felt half baked, the issue might not have been the model. it might have been the harness wrapping it. swap the harness, run the same model again, and watch what changes. hermes agent is the one i recommend to everyone running local. and especially to anyone who almost gave up on it.

@dotta @papercliping Added to the board.

@gcampton @papercliping This is so fun Did you know we have walk cycles for thousands of wizards here? I wonder what it would take to get these guys into the office?? forgottenrunes.com/blog/walkcycles

Agent Pixels (FREE) - A Camera view of your company running inside Paperclip @papercliping featuring @dotta as the CEO github.com/gcampton/Agent… agent-pixels.com

@dotta @papercliping I plan on increasing pixel count. They are a bit basic at the moment. But this is more of the idea

@dotta @papercliping Oh wow, can't believe I never looked at forgottenrunes. Yeah that's 100% doable.

@dotta @Amit_0717 @papercliping ditto, but I can't really take credit for it. Pixel Agents VSCode extension already existed. And paperclip is the best automation tool currently on the market. I just married the two.

@Amit_0717 @gcampton @papercliping I'd love to