Thinking Methods

Name: Thinking Methods
Author: ysskrishna

Quick answers are easy. Good decisions are not.

You ask your AI assistant, "What could go wrong with Friday's launch?" and you get a handful of reasonable-sounding bullets. Fine. Not what you wanted.

What you wanted is what a sharp colleague does. Assume it already failed. Find the specific ways it fails. Rank them. Say who fixes what, and when to pull the plug.

Thinking Methods teaches your assistant to do exactly that. Ask the same question and it runs a proper pre-mortem.

Stop getting answers. Start getting thinking.

Seventeen methods. One for every kind of stuck.

Before you commit. Pre-mortem finds how your plan fails while you can still change it. Critical thinking pokes holes. Ethical thinking asks who could get hurt.
When something keeps breaking. Five Whys digs to the real cause. Systems thinking shows why it keeps coming back. Analytical thinking breaks down why a number moved.
When you have to choose. Tradeoff analysis settles A versus B, build versus buy. Prioritization ranks the backlog and draws the cut line. SWOT and Strategic thinking help you pick a direction.
When you need a number, fast. Fermi estimation gives you a sound ballpark and tells you if the idea is even feasible.
When you need fresh ideas. Creative thinking gives you many options. Lateral thinking breaks you out of ruts.
When you want every angle. Six Thinking Hats keeps facts, gut feel, risks and benefits apart.
When you question the defaults. First-principles thinking rebuilds from the basics. Design thinking starts with what users really need.
When you do not know where to start. The method selector picks one and runs it.

See what a pre-mortem hands you

The README's own usage prompt is a Friday database cutover: "What could go wrong with the database cutover next Friday?" The pre-mortem skill is for before a launch or commitment, when you imagine it failed and find out why. Each skill also ships a worked example in its references folder.

That is one method. You get sixteen more.

You do not have to learn any of it

No names to memorize. Just say what is on your mind: "this keeps breaking", "A or B?", "poke holes in this", "help me think this through". Your assistant picks the right method.

It also knows when to stay out of the way. Every method comes with a note on when not to use it, so you do not get a long ritual for a small question.

Ready to try it?

Works with Claude Code, Cursor, Codex, Gemini CLI and many more. The one-line installer supports 70+ assistants.

Install from GitHub (npx skills add ysskrishna/ai-agent-skills)
npm package for OpenCode, Pi, and other hosts that load skills from npm

Keep exploring

Your AI Agent Shouldn't Start From Zero Every Session: why reusable skills beat pasting the same prompt into every chat.
Teach AI Agents to Think From Multiple Angles with Six Thinking Hats Skill: one method, hat by hat.
Thinking Skills for AI Agents: From Brainstorming to Critical Analysis: the first ten, and when to use each.
AI Agent Skill Validator: writing your own? Check it in your browser, free.

Frequently asked questions

How is this different from just asking my AI to do a pre-mortem?

You can ask, and it will give you an answer. The difference is consistency. A skill follows a structured method in SKILL.md instead of improvising one. Each description lists everyday phrases it fits and when to skip it, so you do not have to write the prompt. It also steps aside when a method does not fit.

How do I stress-test a plan with AI before I launch?

Ask "what could go wrong with this plan?" With the pre-mortem skill installed, your assistant treats it as a pre-launch commitment: imagine it failed and find out why.

How do I get my AI to find the root cause of a failure?

Say "why did this happen?" or "this keeps breaking". The Five Whys skill is for when something failed or keeps recurring and you need the real cause. For problems that keep coming back across teams or services, Systems thinking is the better fit.

Can it help me choose between two options?

Yes. Ask "A or B?" or "build vs buy". The tradeoff skill is for when you must pick between concrete options.

What do I need to use it?

An AI coding assistant such as Claude Code, Cursor, Codex or Gemini CLI. Most hosts use one line from GitHub (`npx skills add ysskrishna/ai-agent-skills`). OpenCode and Pi can install the npm package `@ysskrishna/ai-agent-skills` instead.

Is it free?

Yes. It is open source under the MIT license.

Will it make my decisions for me?

No. It organizes the thinking and shows its reasoning so you can check it. The decision stays with you.