Skill · Think
Pre-mortem: how will this fail?
Run a pre-mortem with AI: imagine your project has already failed, then work back to the causes, warning signs and fixes while there is still time.
The skill#
Act as an experienced project manager running a pre-mortem.
The project: [DESCRIBE YOUR PROJECT]
Goal and deadline: [GOAL AND DEADLINE]
Imagine it is six months from now and the project has clearly failed.
1. Write a short, realistic story of how it failed.
2. List the ten most likely causes, from most to least likely.
3. For each cause, give an early warning sign I could watch for.
4. For the top five causes, give one preventive action I can take now.
5. Note which risks are outside my control and how to limit the damage.
Be specific to this project, not generic.
Return a table: Cause | Likelihood | Warning sign | Preventive actionHow to use it#
- Copy the skill and describe the project, its goal and deadline.
- Read the failure story first; it often surfaces risks lists miss.
- Assign each top risk an owner and put the warning signs in your calendar.
Example#
Describe a website relaunch due in eight weeks. The failure story blames late content and a broken redirect plan; the table lists ten causes with warning signs such as "copy not approved by week 3".
Quick and strict versions#
Use the quick version for everyday checks and the strict version when the result matters.
Imagine this project failed. Give the five most likely reasons, and one thing I can do now to prevent each.
Project: [DESCRIBE YOUR PROJECT]Act as a risk lead running a formal pre-mortem.
Project: [DESCRIBE YOUR PROJECT]
Team and resources: [WHO AND WHAT]
Dependencies: [EXTERNAL PARTIES, SYSTEMS]
1. Assume it failed; write three different failure stories (people, technical, market or external).
2. Extract every distinct cause and score likelihood and impact from 1 to 5.
3. For causes scoring 12 or more, give a warning sign, an owner, a preventive action and a contingency plan.
4. Identify single points of failure.
5. Suggest a review date and what to check then.
Return: Failure stories | Risk register (cause | L | I | score | sign | owner | prevention | contingency) | Single points of failureUse it as an agent skill#
Save this as SKILL.md in a folder named after the skill, and agents that support skills (Claude, Claude Code and other skill-aware agents) can load it automatically when the task fits.
---
name: pre-mortem
description: Use when the user is about to start or launch a project and wants to anticipate how it could fail.
---
# Pre-mortem: how will this fail?
Fill each [BRACKETED] field from the conversation. If one is missing, ask the user for it before starting.
Act as an experienced project manager running a pre-mortem.
The project: [DESCRIBE YOUR PROJECT]
Goal and deadline: [GOAL AND DEADLINE]
Imagine it is six months from now and the project has clearly failed.
1. Write a short, realistic story of how it failed.
2. List the ten most likely causes, from most to least likely.
3. For each cause, give an early warning sign I could watch for.
4. For the top five causes, give one preventive action I can take now.
5. Note which risks are outside my control and how to limit the damage.
Be specific to this project, not generic.
Return a table: Cause | Likelihood | Warning sign | Preventive action
Frequently asked questions
What is a pre-mortem?
A technique where a team imagines a project has already failed and explains why. Because the failure is assumed, people name risks they would otherwise keep quiet about.
When should I run one?
Before you commit time or money: at kickoff, before a launch or before a big decision.
How is it different from a risk list?
Starting from a failure story makes the risks concrete and connected, instead of a generic checklist.