This AI prompt runs a pre-mortem on your project before you start. It imagines the project already failed, ranks the 5 most likely reasons & gives you an early warning sign and one action for this week for each.

Project failure planner

The idea behind this prompt is simple. Before you start a project, you imagine it already failed. Then you figure out why. It’s called a pre-mortem, and it’s a useful exercise. The question is whether this prompt actually gets you there.

Here’s the excerpt:

Role: You're like a super-duper risk checker who knows how to plan stuff. Your whole thing is finding ways projects can go sideways and how to stop it.

Task: Do a "pre-mortem" for this project. Pretend it's already a huge disaster. Figure out the most likely reasons it tanked, what exactly went wrong, and what we can do now to make sure that doesn't happen.

[Project description placeholder, analysis steps, markdown table output format, example project included]

The OP says the structure matters more than the wording. He’s right about that, but maybe not in the way he thinks.

The structure is doing all the work. The role is doing none of it. “Super-duper risk checker who knows how to plan stuff” doesn’t change what the model produces. Remove it entirely and the output is the same.

But there’s a bigger problem.

The prompt asks the model to imagine failure, identify failure points, and develop mitigation strategies. That sounds thorough. It’s actually just 3 ways of saying “what could go wrong and how do we stop it.” There’s no instruction to prioritize. No way to separate a project-killing risk from a minor annoyance. The model will give you 10 rows in a table and treat them all as equally important.

The example project is useful (launching an online coffee business), but it also lets the model pattern-match to generic startup risks instead of thinking about your specific project.

Strip the role. Cut the 3 steps down to 1 real question. Add a prioritization constraint.

The prompt that runs a project pre-mortem

I'm planning this project: [describe it in 2-3 sentences].

Imagine it failed completely 6 months from now. List the 5 most likely reasons it failed, ranked by how likely each one is. For each reason, give me 1 specific early warning sign I'd see in the first 2 weeks, and 1 action I can take this week to prevent it.

Put it in a table: Failure Reason, Early Warning Sign, This Week's Action.

The early warning sign is what makes this version different. The original prompt tells you what could go wrong. This one tells you what to watch for right now.

A prompt works when it asks for something specific.

Original post by u/promptoptimizr. Link

When the plan needs more pressure

Want the plan pushed harder? The grill me prompt interviews you about every part of it. Still deciding whether to start at all? Use the hard decision prompt, or the stress test prompt if it’s still just an idea.

This teardown first ran in issue 1.