Getting started

This is a library of thinking frameworks rebuilt as skills an AI agent can run, and that humans can follow by hand. Each one takes a messy situation and produces a concrete artifact, not a wall of prose: a ranked risk register, a weighted option matrix, an argument map, an honest base-rate estimate.

Who it is for

Deciding or stuck

Anyone facing a real decision or a recurring problem who wants a structured move, not vibes.

Solo operator + AI

Working alongside an agent and want to get unstuck or decide well, right now.

Skeptics

Want to check whether these methods are actually supported by evidence. Spoiler: it is mixed, and we say so.

Builders

Agent and plugin builders who want to call these skills programmatically.

The fastest path: let the advisor choose

Start with the Framework Advisor → Describe your situation in plain language and get a Thinking Plan: the one or two frameworks worth running, in order, why each fits, and what to skip. It is the front door to everything else.

Your first framework: a premortem

A premortem Evidence: S/M stress-tests a plan by imagining it has already failed and working backward to why, then turning each cause into a tripwire and a pre-decided response.

1. Pick a real decision you are about to commit to: a launch, a hire, a migration.

2. Run the skill in Claude Code (or follow its numbered steps by hand with a pen):

   /think-premortem "we're about to launch a free tier to drive signups"

3. Get a ranked risk register. For each top risk: a leading signal, a mitigation, an owner, and a kill criterion.

That artifact is the point

The deliverable is the register of what could go wrong and what you will do about each, not a vague feeling of caution. Every skill in the library ends in a concrete artifact like this.

How to read a framework page

Every framework page is layered so you can stop at any depth: a 20-second quick-facts card, the mechanism and procedure, a worked example, then the full evidence dossier with graded sources.

How to read a page The four layers, and how to pick your depth.

The evidence model What the S/M/P/V/A/C/X tiers mean, and why honest grading is the product.

You do not need an agent

Each framework is a procedure you can run by hand with the template on its page. The agent just makes the mechanism cheap to run and enforces the structure.