Open the guide
The Blade
A six-question protocol for thinking clearly about any decision — fast. Ends in one action and one kill, not a list of options.
What am I missing, and what should I actually do?
- Status
- working
- Reviewed
- Jul 31, 2026
What The Blade is for
Most decisions get stuck in one of two places: you have too many options and can’t choose, or you have a plan but something feels off and you can’t name it.
The Blade is a fast-path thinking protocol. Six questions. One screen of output. You walk out with one thing to do next and one thing to stop doing.
It is not a brainstorm. It does not produce a list of considerations. A recommendation without a kill at the end is not a decision — it is a wish with structure.
Run it on any decision where you feel unclear, overconfident, or stuck.
The six questions
Work through these in order. Write your answers — even rough notes. The act of writing forces the precision that thinking out loud avoids.
1. State
Name the psychological state driving this decision right now.
Not the reasoning — the emotional charge underneath it. Fear of missing out. Attachment to work already done. Urgency that appeared this week. Ambition that wants a plan. Fatigue that wants to stop thinking.
Write it in one sentence.
If the charge is high, the analysis that follows will tend to serve the state rather than the truth. Naming it does not remove it. It tells you to hold the conclusions more lightly.
Write: The state driving this is _______.
2. Failure
What would guarantee the opposite of what you want?
Do not list risks. Invert the goal completely. If the goal is to ship this successfully, ask: what would guarantee it fails publicly and damages trust? If the goal is to exit this situation well, ask: what would guarantee you exit badly and regret it?
List the top two or three failure conditions seriously. Do not dismiss them as unlikely. Then ask: are any of these conditions currently present?
Write: The things that would guarantee failure are _______. Currently present: yes / no / partially.
3. Unseen
What information, person, or fact is not in the room that would change the framing?
Your current framing is built from what you can see. System 1 fills the gaps with plausible assumptions and produces confidence anyway. The more coherent your story feels, the more important this question is.
Ask: who is not in this conversation? What data are you not looking at? What time horizon are you ignoring? What would a person with the opposite view say is obviously missing?
Write: What is not in the room: _______.
4. Actors
Classify everyone between you and the outcome.
For each person or group that matters:
- Buyer — owns the problem, lives with the outcome, would pay to solve it
- Gatekeeper — controls access or approval, but does not own the outcome; often has something to lose if you succeed
- Amplifier — gains status or credibility by sharing your success
A strategy that pitches gatekeepers as if they are buyers fails structurally, regardless of how well it is executed. If the person you are trying to reach is a gatekeeper, the question is not how to improve the pitch — it is how to reach the buyer instead.
Write: Buyers: _______. Gatekeepers: _______. Amplifiers: _______.
5. Constraint
What single bottleneck, if doubled, would most change the outcome?
Not the most urgent thing. Not the most visible problem. The one thing that, if it were twice as good or twice as fast, would pull everything else forward.
Everything that is not the constraint is noise — improving it produces zero system improvement. Name the real constraint. Anything you’re doing that doesn’t serve it directly should be questioned.
Write: The real constraint is _______. What I’m doing that doesn’t serve it: _______.
6. Cut
End with exactly one action and one kill.
Not a priority list. Not a set of considerations. One thing to do next — the specific move that follows from what the previous five questions revealed. And one thing to stop doing — the activity, assumption, or effort that is consuming resources without serving the constraint.
If you cannot name a kill, you have not made a decision. You have produced an analysis.
Write: One thing to do: _______. One thing to stop: _______.
How to read the output
After completing all six, read your answers together. Look for:
- Does the state (Q1) explain why any of the failure conditions (Q2) feel unlikely? If so, the state is protecting you from seeing them.
- Does the unseen information (Q3) change who the real buyer is (Q4)?
- Does the constraint (Q5) match what you actually spend time on? If not, realign before choosing the action (Q6).
The questions are not a checklist. They are a search for the thing you would not have named without being asked.
What this guide does not do
The Blade surfaces the structure of a decision. It does not tell you what to decide. The final judgment belongs to you — not to the analysis.
If two runs of The Blade on the same problem reach different conclusions, that disagreement is a signal, not an error. It means the problem is more complex than one fast pass can settle. Run a deeper session with the full framework set, or wait for information that would break the tie.
The Blade is the fast-path version of a larger critical thinking skill with 8 framework sets and 46 individual frameworks. Use it as the default. Escalate to the full protocol when the Blade’s answer feels thin or when two runs disagree.
Open the instrument