scaled-thinking is a family of free, open modules for your own AI agent. Each module is a specialist (an agent, a service or a capability) that extends what you and your agent can do. Take one or several: most modules run only while they are needed.
For anyone who uses AI and wants to grow with it, not depend on it.
Say you have a talk next week. The AI asks what your audience should remember, suggests a strong opening and builds a quick live poll. It says why each idea works; you choose what to keep and take the method with you.
Copy this into your AI chat. It reads the list of modules and suggests what fits you. If it cannot open the link, use a chat with web search.
Read https://raw.githubusercontent.com/scaled-thinking/scaled-thinking/main/MODULES.md,
the list of scaled-thinking modules. Based on what you know about me and my work,
suggest the modules that would help me most and say why. If you know too little about me,
ask me three short questions first. I decide what to try; don't set anything up yet.
Reply in my language.
- present-together: Give better talks: an AI coaches you and builds the interactive parts.
- open-research-loops: Test your research idea with AI and see what really holds up.
The list also shows the modules still in progress. Every module is free and runs on the AI you already have; the list says when one needs more. What you share goes only to the AI and services you choose. scaled-thinking has no server, so it never sees it. Each module is yours to keep: you may use, change and share it.
Most AI tools simply hand you the result. That's handy, but you learn little and need the tool again next time. Competence-enhancing AI does the work too, but takes you along, with learning methods from research:
- It asks before it tells (Socratic dialogue). A question, then a small hint; the answer comes last.
- You guess first (predict, then reveal). You commit to a guess, then see what really happens.
- It challenges you. It looks for gaps and weak points in your idea; you decide what to do.
- You say it in your own words. Explaining shows what you have really understood.
- You practise, and it checks later (spaced repetition). A week later a short test shows what stayed.
It adapts to you. It first finds out what you already know, then compares how sure you feel with how you actually do, and adjusts: more help when you are new, less as you get better. You confirm every change, and your profile stays on your computer.
Every method has a name, so you can use it again without the AI. In a hurry? Say "just do it".
Competence-enhancing AI makes you more capable, not more dependent: it explains, you decide, and you keep what you learn.
Your AI does the work; you decide. Every module follows four rules:
- Check, don't trust. Every result comes with what you need to check it.
- A second opinion. Tests or a different AI check the work, never the AI that did it.
- Your limits. You set them, and you can stop and go back at any time.
- Nothing hidden. Whatever nobody has checked yet is marked.
Soon AI may do almost any task better than any of us. What stays with people is deciding what is worth doing, and answering for it. That takes competence, so I let AI do the work and help the people who direct it grow.
Everything here is free and open, and it grows with you: bring your ideas, your feedback or a module of your own. Now that AI lets anyone shape software, software should be there for everyone.
Questions and feedback: contact@scaled-thinking.com


