Four instruments that break your own reasoning, in your browser. Nothing here uses an AI, nothing leaves your machine, and every verdict is computed, not guessed.
Socratic Interrogator
State a belief. The interrogator asks why, then asks why again, until your justifications run out — then it shows you what the structure actually looks like: circular, contradictory, or resting on an unexamined assumption. It never tells you whether you are right; it only examines whether your reasoning holds together.
Fallacy Foundry
Paste an argument. The foundry scans it for the shapes of twelve classic fallacies — cue phrases and argument skeletons — and flags them inline, with a note on what to check. It is a detector, not a judge: your text never leaves this page, and the final call is yours.
Self-Reference Lab
Build a small universe of sentences that talk about each other's truth, and watch the engine try to decide what is true. Some systems settle; some never can. The engine uses the same fixed-point semantics logicians use for the liar paradox — and it will happily show you the exact sentence (or chain of sentences) that breaks your universe.
Preference Paradox
Rank things the honest way — one head-to-head at a time. When you finish, the lab builds your preference graph and checks whether it contradicts itself. It can. You can genuinely prefer A over B, B over C, and C over A, all at once. Most people find out about their own taste cycles this way.