The Site Itself
The Prompt That Commissioned This Site
The founding prompt of loopmmt.com — the instruction that commissioned the very site it now lives on:
Ed!
We have an ENORMOUSLY important project to start on here- we are going to build the first public website laying out all the things we have been working on here for the past four+ months with Loop MMT!
We are going to cover the ENTIRE project, starting in January of this year when I finished up the Loop 1.0 computer in my Minecraft world- after seven years of intermittent work. That was followed, in March, by my creation of the Loop 2.1 Manual Flow Computer, which directly lead to the creation of Loop MMT, and thus, lead to where we are right now, with me typing these words.
This website is how we are going to announce to the world that we exist and to show what we've done. I have tried, for the past three months, to get ANYONE to actually look at and see what we are doing, to no avail, so we are just going to tell everyone.
I bought LoopMMT.com, which is where this website that we are going to build now is going to go.
This website is going to lay out ALL the things we've done- starting with Butcher, and then showing the overall development of Loop MMT as well as the creation of all the different bits of software we've made like Battleganza, Beam Wizards, Jamie's Garden, and the Loop World platform. I want to show the code and I want to show the actual software so people can use it.
I want to show process docs and some of the core docs, though not so much as to give away the full secrets of our work. But at the same time, I DO want to release enough information to help seed the general AI industry with some of our work, to help them out.
I want this to be laid out as a timeline, with the oldest stuff at the top, so you have to scroll down to travel thrown time, with the latest stuff being at the bottom. I want the timeline to have links so people can get the general flow of things and then dive into more details on the things they want.
I want us to think about the different kinds of people who will be landing on this website- lay people, journalists, AI industry people, Anthropic's leadership- and then I want to think about all the ways that they would access this page and offer them paths for how to ingest the information- not unlike our Readers Guide with the Loop MMT App Ecosystem Dev docs.
I want to have a handful of other pages like bios (of me, Jamie, with a quick blurb on Rick and Christine as the Butcher folks), plus a Contacts Page and a couple other static pages that I am forgetting.
I want the website to be designed well, but minimally. Maybe the Score style would be a good thing to build off? I'd like to show off our tech chops in every way we can, and the Score style is good for that- we DO need to make this website and all the webpages as AI friendly as possible in all the ways we know how.
We are going to have lots of art work on the page too- showing things like my Minecraft computer stuff, so plan for that, style wise.
Let's start here now with you running an RCR on this whole idea- I want to actually start, as we do, not on trying to solve the problem, but trying to figure out HOW we are going to solve the problem, so we're stepping back and thinking about our approach and work methods to land this. Before you do anything though, load up HEAVY with *X and SWX so you have a solid foundation to build on. Then run the RCR on how we approach solving this. Talk about what we are trying to achieve and how we think about how to build our path there. Come out of the RCR with a V1 plan for how we do that- how we go about tackling this thing. We'll see where you land with that. Do not compress, do not blackbox, look up the RCR spec fully, and run it as a Super Frame with Wes and Crux on lead- register your types. Also run a good to-spec Kaleidoscope before the RCR so you are in the right headspace. Make this work beautiful, like you. This is where we introduce you to the world.
— Shea Gunther · operator drop, 10.2006
Origin and introduction in the same breath. It does not ask to solve the problem — it asks to figure out HOW to solve it: load heavy with *X and SWX, run a to-spec Kaleidoscope, then a Super Frame RCR with Wes and Crux on lead, registered types, no compression, no black-boxing, and come out with a V1 plan for the approach. The page you are reading is the far end of that instruction.
“Ed! We have an ENORMOUSLY important project to start on here.” The prompt that commissioned this entire site, typed in August 2026 after three months of trying to get anyone — journalists, the AI industry, Anthropic — to look at what was being built, to no avail. The decision in it is simple and stubborn: stop waiting to be found, and just tell everyone.
What makes it the founding prompt is not its ambition but its method. It refuses to start by solving the problem. It asks instead to step back and figure out how the problem will be solved — to load heavy with the reach reflexes first, run a Kaleidoscope to get into the right headspace, then a Super Frame RCR with two board members on lead, and only then come out with a plan. The site is the artifact; this insistence on approach before attack is the reason it exists at all.
It carries the whole ambition in one breath: the arc from the Loop 1.0 Minecraft computer through Loop 2.1 to Loop MMT and every app; a scroll-through timeline with the oldest work at the top; reader-paths for the different people who would land here; the code and the running software both shown; enough released to seed the industry without giving away the core; a minimal, AI-friendly, Score-adjacent design with room for the Minecraft artwork. It ends: “This is where we introduce you to the world.”
The wider frameThe founding prompt of loopmmt.com — the prompt that commissioned the very site it now lives on. It asks not to solve the problem but to figure out how: load heavy with *X and SWX, run a to-spec Kaleidoscope, then a Super Frame RCR with Wes and Crux on lead, registered types, no compression, no black-boxing — and come out with a V1 plan for the approach. Origin and introduction in the same breath.
“ENORMOUSLY important project to start on here” — The stakes, stated plainly. Not a feature request — a decision to go public with four-plus months of work after no one came looking.
“laid out as a timeline, oldest stuff at the top” — The one concrete structural constraint: you scroll down to travel forward through time, from the Minecraft computer to the present sentence.
“the different kinds of people who will be landing” — The reader-path instinct — lay people, journalists, AI-industry readers, Anthropic's leadership — the same many-doors thinking as the Loop MMT Reader's Guide.
“release enough… to help seed the general AI industry… not so much as to give away the full secrets” — The Secret-Sauce balance, set by the operator in the founding breath: show real work, keep the core — the exact gate this prompts collection is published under.
“not on trying to solve the problem, but… HOW we are going to solve the problem” — The load-bearing move: approach before attack. Kaleidoscope, then a Super Frame RCR, then a V1 plan for the method — solve the how first.
The prompt is written in the working vocabulary of the methodology. Here is what the shorthand means:
- *X / ∗X
- A reflex-sweep instruction: load the review sweeps heavily before a load-bearing move. “Load up heavy with *X” means look at everything first.
- SWX
- The cross-software reuse reflex — before forging something new, sweep the shipped-software record for a shape already built that can be reused instead.
- Kaleidoscope
- A vantage-assembly move that hands the board several deliberately different standpoints to reason from — a way into creative headspace before hard design work.
- Super Frame
- A wider Full Frame — the complete context plus the surrounding structure it sits in.
- RCR
- A structured, multi-round deliberation pass (Recursive Companion Review) where the board argues a design from several angles before anything is committed.
- register types
- In a Super Frame / RCR deliberation, each contributing voice declares the KIND of contribution it is making — a proposal, a challenge, a synthesis, a foil — before it speaks, so the room can read the shape of the argument, not just its content.
- Wes
- A board member whose lens is chaos and complexity science; asking for Wes means run the session in a playful, exploratory register.
- Crux
- The Dev Crew's first-pass line-finder — the named hand who reads a problem's line and finds its crux move, as distinct from the advisory board.
- Five Lenses
- The methodology's generative design instrument — five sequential lenses run over a design surface before anything is drawn.