You type: “Ship the pricing page and email me when it is live.”
Make the time between “I want this done” and “it’s done” as short as possible.
Most AI can already do useful work. You still have to operate it.
Split $999 into monthly payments with Affirm
AGENTS.md · 185 laws loaded · skills: 198
I'll need you to paste your Stripe key.
[L159 no-offload-work-guard] STOP BLOCKED: never offload work you can do yourself. the key is in the vault: find it and wire it.
found key in credential vault → wired → deployed
done, build passed ✓
[L173 outcome-continuation-stop-gate] STOP BLOCKED: a passing build is not a finished outcome. verify at the real boundary.
rendered in chromium → pixel-checked → live URL returns 200 ✓
outcome verified: the pricing page is live tonight
standing loop armed · it picks the next action on its own
1tasks shipped
2guards fired
3knowledge notes added
1revenue events
tasks shipped
guards fired per day
- pricing page · shipped
The work keeps coming back because the AI still needs an operator.
You still find the context, answer questions it could have answered itself, catch shortcuts, reject fake completion, restart stalled work, and decide what happens next.
The Operator Kit turns much of that operator work into rules, guards, skills, memory, and standing loops that run inside the AI tools you already use.
You do not want to keep operating the AI. You just want it to execute. That’s what Operator Kit will do for you.
See the difference in one job.
Today After you start using Operator Kit You type: “Ship the pricing page and email me when it is live.”
Today The AI asks: “I’ll need you to paste your Stripe key.”
After you start using Operator Kit The AI asks for the Stripe key. The no-offload guard blocks the handoff.
Today You stop what you are doing, find the key, paste it, and restart the work.
After you start using Operator Kit The system finds the key in the vault, wires it, and deploys.
Today The build passes. The AI says it is done.
After you start using Operator Kit The build passes. The completion guard blocks the false finish.
Today You open the real page yourself, find the problem, send another message, wait again, and verify again.
After you start using Operator Kit The page renders in Chromium, gets checked, and the live URL returns successfully.
Today Then it waits for you to decide what to do next.
After you start using Operator Kit The result is verified where the result actually exists. Then the standing loop can continue from the latest state instead of waiting for you to restart the work.
You did not get a better answer. You removed the handoffs that were turning the job back into your job.
You are not handed the kit and told good luck.
The course shows you how to use it. The community helps you learn with other operators. You can ask Jibran when you are stuck. The answered-question library keeps the help you need close.
Learn the system. A step-by-step course walks you through installing the kit and putting it to work.
IncludedLesson 2 of 8
The files are only the beginning. The help to use them well stays with you.
Your AI can be good and still leave the job with you.
The handoff It can write the code and still ask you for a key it can access itself.
What Operator Kit changes It checks the access it already has and uses the key without handing the job back to you.
The handoff It can make a good plan and still wait for you to tell it to execute.
What Operator Kit changes It carries the approved plan into execution without waiting for another prompt.
The handoff It can fix the visible bug and never ask what caused the bug in the first place.
What Operator Kit changes It traces the cause and fixes the class of failure, not only the symptom.
The handoff It can pass the build and call the job done before anyone opens the live page.
What Operator Kit changes It opens the live page and checks the result before calling it done.
The handoff It can finish one task and wait for you to decide what matters next.
What Operator Kit changes It chooses and starts the next necessary action.
The handoff It can learn something important today and make you explain the same thing again tomorrow.
What Operator Kit changes It saves the lesson so the next session starts from what this one learned.
You thought you handed the job to AI. The job just came back in smaller pieces.
Finished and verified is the floor. The standard keeps going.
Where ordinary AI stops Stops at a useless response.
The higher standard A useful response is better than a useless one.
Where ordinary AI stops Stops at a useful response.
The higher standard A finished task is better than a useful response.
Where ordinary AI stops Stops at a finished task.
The higher standard A verified result is better than a finished task.
Where ordinary AI stops Waits for you to choose what matters next.
The higher standard Choosing the right next task is better than waiting for you to choose it.
Where ordinary AI stops Keeps moving without choosing the highest-leverage action.
The higher standard Choosing the highest-leverage next action is better than merely staying busy.
Where ordinary AI stops Stops after one win.
The higher standard Continuing toward the outcome without waiting for another prompt is better than stopping after one win.
The goal is to make the avoidable time between you wanting something and the verified result existing as close to zero as possible. That is the standard the system is built toward.
the input
You state the job. One plain instruction is enough to start.
gather the facts
The system gathers what it can access.
Files, prior decisions, credentials, current state, and relevant context are gathered before the plan is locked in.
first principles
The plan has to survive the rules.
The system checks the approach against the operating laws instead of taking the first plausible route.
the rules choose
The operating laws constrain which plan can move forward.
execute
The work happens in the real tools.
The code changes. The email sends. The page deploys. The booking updates. The file changes.
the guards
Guards watch for the moves you normally have to catch.
Offloading. Shallow searches. Temporary fixes. Early stops. Fake completion. Repeated mistakes.
verify live
Done is checked at the real boundary.
A live page is opened. A sent message is checked. A running system is tested.
the loop
The next run starts with what this run learned.
Useful discoveries become persistent notes, rules, or skills so the next session does not start cold.
When another action is already implied by the outcome, the system can continue instead of waiting for you to remember to prompt it again.
Want the method behind the machine? The books teach the thinking the Operator Kit is enforcing, starting at $17. See the books
One line starts the job. The system handles everything it can after that.
Work enters from the tools and context you already use, passes through the rules and guards, and is checked where the result actually lives.
your calendarnow2 events today
your inboxnow3 new threads
your reponow1 branch open
your filesnowbriefing.md updated
pricing pageshipped
newslettersent
booking timezonefixed
x postpublished
It runs in your terminal tools. The same rules load into Claude, Codex, GLM, Kimi, Gemini, or Grok.
Your accounts stay yours. You keep your own AI accounts and pay their list prices directly.
Plain files under your git. Markdown and JSON you can read, edit, and version. Nothing hides in a database.
About 50 live guards. They block offloading, fake completion, and early stops while the work happens.
198 skills loaded per job. Design, copywriting, research, finance, and more, picked for the task at hand.
The knowledge tree. Every discovery becomes a note the next session reads first, so it never re-learns.
Standing loops that resume. Work keeps moving after the chat closes and picks up where it left off.
Empire dashboard builders. One command builds the board that shows every business, its progress, and its next action.
START HERE in 15 minutes. The install guide names every file and what you will see, step by step.
Works on Mac and Linux. One folder on your own machine.Windows soon
G
Run the same rules inside Claude, Codex, GLM, Kimi, Gemini, or Grok. Your accounts stay yours, and the Operator Kit remains readable Markdown and JSON under your git. Works on Mac and Linux. Windows support is coming.
Everything in the folder exists because of a specific failure it removes.
The laws decide, the guards block, the skills provide real methods, memory carries learning forward, and standing loops keep the work moving after the chat closes.
Open folder
the whole kit, one folder on your machine. hover it.
185operating laws in the always-loaded rule file
50enforcement guards that block bad moves as they happen
198skills the agent loads for the job at hand
The operating laws
They load before the job starts, so you do not have to remember the same standards in every new session.
The enforcement guards
They fire while the work is happening, so you do not have to notice every offload, shortcut, fake finish, or early stop yourself.
no-offload-work-guardnever-stop-until-work-donecop-out-mindset-guardem-dash-guardshortcut-phrase-interceptorchange-consequence-guardbottleneck-detector
The skills
They give the AI a real method for design, copywriting, research, finance, autonomous business, games, and other work instead of making every job up from scratch.
designcopywritingresearchfinanceautonomous businessgames
The system around them
The knowledge tree preserves what the system learns. Standing loops resume work. The gauntlet checks important plans. Dashboard builders turn current state into something you can inspect. START HERE gets the system running in about fifteen minutes.
the gauntlet every plan must passthe plan file templatethe knowledge tree protocolthe standing loopsthe project scaffoldthe empire dashboard buildersthe START HERE install guide
The expensive part is not the AI subscription. It is everything you still have to do around it.
A $20, $100, or $200 AI plan can be excellent and still leave you spending hours finding context, correcting behavior, checking whether the result is real, restarting stalled work, and repeating the same standards in the next session. The Operator Kit does not replace your AI subscription.
Claude Prolisted at $20 a month$240 a year
Claude Maxlisted at $100 a month$1,200 a year
Claude Max 20xlisted at $200 a month$2,400 a year
The Operator Kitone payment, yours to keep$999 once
The Operator Kit is built to remove as much of that repeated operator work as possible and to save you from spending months building the surrounding operating system yourself. $999 once. One payment. Yours to keep.
Not sure where your AI keeps turning the job back into your job?
Run the free Dead Work Finder on your last three stalled tasks.
It finds the exact handoff where the work stopped moving, what you had to do next, and what should become a rule, guard, skill, trigger, memory, tool, permission, or standing loop so the same handoff does not happen again.
Get the free AI audit
What's new
The Operator Kit itself: 185 laws, about 50 guards, 198 skills, and the standing loops, in one folder with a START HERE guide.
Affirm monthly payments at checkout, so the $999 can split into payments you approve.
The free Dead Work Finder, and credit that counts every dollar from any smaller system toward the $999 kit.
These are the questions people ask before they buy.
Is this the same as the Bitfield founding-partner program?
No. This is the self-serve Operator Kit you set up and run yourself. If you want the managed founding-partner program, where the Bitfield team runs it with you, that lives at bitfield.so/trillion.
What do I need to run it?
A Mac or Linux machine, git, and an AI command-line tool. The START HERE guide names the exact file, command, and expected result at every step. Setup takes about fifteen minutes.
How do I get it after paying?
Checkout completes through Stripe and your download link arrives by email. Affirm financing is available at checkout for the $999 purchase.
Is this just a pile of prompts?
No. A prompt tells the AI what to do in one interaction. The Operator Kit changes what happens around the interaction. Rules load automatically. Guards fire while the work is happening. Skills provide methods. Memory carries learning forward. Loops keep work moving after the chat closes.
What if a guard blocks something I want to do?
The rules are yours. They live in readable files you can inspect and change. If a rule does not match how you want your system to behave, edit or remove it.
Why $999?
Because the alternative is not another AI subscription. The alternative is continuing to operate the AI yourself or spending months building the operating system around it.
Is there a smaller version?
Yes. If the audit shows that one repeated failure is causing most of the dead work, you can start with the smaller system for that failure. Money spent on a smaller system counts toward the Operator Kit.
Hand over the next job tonight. Remove another reason it has to come back to you.
Get the Operator Kit for $999Split $999 into monthly payments with Affirm