AI command system · Claude & ChatGPT
Stop asking AI questions. Start switching its mode.
Most people use AI like a vending machine: question in, average answer out. The Practitioner's Playbook is a command system that puts Claude or ChatGPT into distinct reasoning modes, so the work you hand over is right, not just right-looking.
- 16 commands
- 6 research disciplines
- 6 acceleration strategies
- Works on free plans
you › Assess this acquisition thesis.
Looks strong. Growth is attractive and the market is large...
you › /brutal
Ranked by damage: 1) revenue concentration in two clients. 2) The margin assumption ignores churn. 3) No evidence for the exit multiple.
you › /steelman
The problem
Average questions get average answers.
You are accountable for the work. The model is not. Without a system, you get fluent output you still have to distrust, check and rewrite.
Without a system
Long prompts retyped every time. Confident hallucinations. Padding, hedging and recaps. Every session starts from zero.
With the Playbook
One short command sets the posture: terse, adversarial, exhaustive, or plain-language. Sources are checked, scope is bounded and "done" is defined upfront.
What's inside
A complete operating system, not a prompt list.
Copy, paste, activate. Every command is written out in full, with installation guides for both Claude and ChatGPT.
Mode-switching commands
From /compact to /ultraplan. Each one changes how the model reasons, not just how it sounds.
Research disciplines
Context loading, output control, anti-hallucination, delta commands, reasoning mode and role stacking.
Acceleration strategies
Front-load clarity, draft in parallel, compress, red-team your own work, bank templates, switch modes.
A taste of the command spec
/compactEssential signal only. No preamble./btwInject side context mid-task without a restart./eli5Plain language, concrete analogies./ultrathinkMaximum-depth, deliberate reasoning./ultraplanFull plan and risks before any execution./brutalHostile-expert critique, ranked by damage./steelmanThe strongest case against your position./insightsNon-obvious patterns and second-order effects.
Plus the rest of the 16 in the full spec.
What you get
Three deliverables, ready to use today.
- The complete command specCopy, paste and activate all 16 commands immediately.
- AI in Practice: the full interactive guideCheat codes, acceleration strategies, practitioner mental models and a 15-minute path from blank page to shipped work.
- AI Infrastructure ResearchThe proof of concept, built using this system.
Proof it works
The system built its own case study.
The Moonshots AI Infrastructure Research Platform covers 120+ positions across the entire AI supply chain, with live web search and institutional-quality analysis on every position. It was built with this system, and the research document included here shows the output.
positions researched
system behind all of it
one-off prompts to hunt down
Built for you if you are
- A consultant, analyst or researcher
- An AI practitioner using Claude or ChatGPT professionally
- Producing work you are accountable for
Not for you if you want
- A beginner's introduction to AI
- A bag of copy-paste prompts
- Something untested on real work
Questions
Before you buy.
Does it work with ChatGPT?
It is primarily built for Claude. The core commands are tested on ChatGPT (GPT-4o), and installation guides for both are included.
Do I need a paid plan?
No. It is compatible with all plans, including free tiers.
Is this a prompt pack?
No. A prompt list gives you text to paste. This is a system of commands, disciplines and strategies that work together, built by a practitioner and tested on real work.
What is the refund policy?
No refunds are offered, so check the "Who it's for" section above before you buy.
The Practitioner's Playbook
Make your AI work like a professional.
One-time payment. Instant access. No refunds.