Pre-mortem
Imagining a failure that has already happened unlocks objections that "what could go wrong?" never surfaces. The framing does the work.
The printable canvas and the handoff into Claude Code are part of a paid plan. See what a plan includes. It needs the Claude desktop app on this machine, and your team and the prompt library already installed in that project — we cannot see your disk, so open it there.
What it is
Gary Klein's technique: before committing to a decision, assume it has already failed, and explain why. The tense is the whole mechanism. Asking "what could go wrong?" invites hedged, socially safe answers; asking people to account for a failure that has definitively happened gives them permission to say the thing they would otherwise keep to themselves.
- The premise
- Stated as history, not prediction. "It is three years from now. This failed badly." Not "might fail" — did fail. Certainty is what unlocks the honesty.
- Silent generation
- Everyone writes their own account before anyone speaks. Without this the first voice anchors the room and you surface one failure mode instead of eight.
- Collection without debate
- All accounts read out, none argued with. Arguing about plausibility is precisely the reflex the exercise exists to suspend.
- Ranking
- Only now do you judge: each failure mode weighed for likelihood and for damage. You are hunting the two or three that are both.
- Warning signs and changes
- For each top failure, what you would have observed in year one — and what you change in the plan today. A pre-mortem that ends in a list has done half its job.
How you run it
- Set the scene in the past tense"It is three years from now. This failed badly." Not "might fail" — did fail. The certainty is what frees people to speak.
- Write silently firstEveryone writes their own account before anyone speaks, or the first voice anchors the room and you get one failure mode instead of eight.
- Read them out without debateCollect all of them. Arguing about plausibility is exactly the reflex the exercise exists to suspend.
- Rank by plausibility times damageNow you can judge. You want the two or three that are both likely and severe.
- Convert to warning signs and changesWhat would we have seen in year one, and what do we change today? A pre-mortem that ends in a list has done half its job.
The prompt
Run this tool in your own Claude
The short prompt starts your partner against the library on your disk. The long one carries everything with it and needs nothing installed.
Your playbook
It lands in the earliest stage this tool suits. Move it on the Playbook page.
You’ll need
- A decision nearly made but not announced
- The people who would have to deliver it
- Genuine permission to say it will fail
You’ll end up with
- A ranked list of failure modes
- The year-one warning signs for each
- Two or three changes to the plan, made now