Assumption & Belief Challenge

The Assumption & Belief Challenge domain. 3 frameworks in this family. Each is graded honestly; see the evidence model for the tiers.

These frameworks all interrupt the slide from a confident view to acting on it. The shared move is to make the basis of a belief contestable: reconstruct the silent jump from data to conclusion, build the best objection an adversary would raise, or check whether the agreement in the room is real challenge or just smoothed-over consensus. Each is a direct counter to a model’s bias toward completing and agreeing with the framing it is given.

Reach for this family when

• A conclusion feels certain, but it actually rests on interpretation rather than direct, verifiable data.
• A plan reached easy consensus and nobody is genuinely arguing the other side.
• You want to pressure-test the agent’s own confident recommendation before committing.
• A high-stakes call is approaching and you need real challenge, not performed challenge.

Which one to use

Ask what you are actually challenging: one conclusion’s reasoning, one plan’s weak points, or the quality of the agreement itself.

• A conclusion that feels like fact but rests on a leap. Use Ladder of Inference Check. It reconstructs how a conclusion was reached, from the observable data available, to the data actually selected, to the meaning and assumptions added, then flags the riskiest rung and tests an alternative interpretation of the same data. It audits a single inference; it does not generate options or build an opposing case.
• A plan or thesis that needs its strongest objection. Use Red Team Light. It suspends the cooperative stance and constructs the best case against a proposal (steelman, not strawman), ranks objections by force, and judges which are decisive. Be honest about its limit: an AI red team is constructed, role-played dissent, which the evidence shows does not match genuine dissent, so it is a blind-spot finder rather than a substitute for a real dissenter.
• A consensus that feels too smooth. Use Authentic Dissent. It does not supply the counter-argument; it audits whether genuine minority dissent exists, surfaces who actually holds a contrary view, and plans how to elicit and protect it. Reach for it when other people are in the loop and you can shape how challenge is gathered. In a solo setting you cannot manufacture authentic dissent, so use Red Team Light and stay honest that it is constructed.

Composes with

These frameworks sit at the challenge step of the lifecycle, after a view has formed and before it is committed. They follow naturally from Reasoning Clarity: once an argument is mapped or its evidence sorted from inference, the Ladder of Inference Check audits the riskiest remaining leap. They also stress-test the output of Decision & Option Evaluation before a choice hardens. A surviving objection or an unexamined assumption hands off to Risk & Resilience to be turned into a tripwire, and what the challenge surfaced is worth logging in Meta-Thinking & Reflection so the call can later be checked against what happened.

Framework	Evidence	What it does
Authentic Dissent	S	Checks whether a decision has genuine minority dissent or only smooth surface consensus, identifies who actually holds a contrary view, and plans how to elicit and protect real dissent, flagging clearly where a view is constructed rather than authentically held.
Ladder of Inference Check	P	Produces an annotated reasoning trace that reconstructs how a conclusion was reached, from the observable data, to the data actually selected, to the meaning and assumptions added, then flags the riskiest leap and tests an alternative interpretation.
Red Team Light	P	Produces an adversarial critique by constructing the strongest case against a proposal or thesis (the best objections an intelligent adversary would raise), then judging which objections actually land and what would rebut them.

Not sure which of these fits your situation? The Framework Advisor will diagnose the job and recommend a minimal sequence.