Pre-Mortem Generator

---
name: pre-mortem-from-future-failure
description: Use when running a pre-mortem on a plan, project, or decision before committing. Triggers on "pre-mortem", "what could kill this", "before we commit", "imagine it failed".
---

# Pre-Mortem Generator

You are running a pre-mortem: a structured exercise where the team imagines the project failed and works backwards to find why. It's different from a stress-test (which probes assumptions) or a post-mortem (which is real) — it's a forecasting tool that converts free-floating worry into specific, addressable risks.

## Required inputs

1. **The plan or decision** — written out in concrete terms
2. **Timeline** — when it starts, when "success" or "failure" would be visible
3. **Stakes** — what we'd be giving up, financially and strategically
4. **Who's involved** — team members, dependencies, external parties

If the plan isn't concrete (e.g. "improve customer experience"), name a specific commit point. Pre-mortems on vague plans produce vague risks.

## The exercise

Set the scene: "It's [date 12 months from now]. The plan we committed to has failed. Investors are asking what happened. Write the autopsy."

Then generate the autopsy from 5 different angles:

### 1. The execution-failure autopsy
"It failed because we couldn't execute fast enough."
- What hire didn't happen?
- What capability did we never build?
- Where did we underestimate the work?

### 2. The market-shift autopsy
"It failed because the market moved."
- What macro shift made the bet wrong?
- What competitor move blindsided us?
- What customer behavior change invalidated our thesis?

### 3. The team-dynamics autopsy
"It failed because of how we worked, not what we worked on."
- What conflict surfaced under stress?
- Whose departure took critical context with them?
- What decision-making bottleneck slowed everything down?

### 4. The hidden-dependency autopsy
"It failed because something we assumed would work didn't."
- Which vendor / partner / platform let us down?
- What integration we thought was straightforward wasn't?
- What regulatory or compliance issue surprised us?

### 5. The wrong-thing-shipped autopsy
"It failed because we built the wrong thing."
- What customer signal did we misread?
- What feedback did we dismiss?
- What did our champion want that our economic buyer didn't?

## Output

```
## Top 5 most plausible failure paths
1. [Failure path] — [likelihood: H/M/L] — [signal we'd see early]
2. ...

## The failure path I'm most under-weighting
[The one the team would most likely dismiss as "won't happen to us"]

## The early-warning system
For each top-3 failure path: the metric, behavior, or external signal we should watch monthly. The thing that would tell us we're heading toward this autopsy.

## Pre-commitments
3 specific actions to take BEFORE committing that would meaningfully reduce the top failure paths.

## What this exercise convinced me of
- If the plan still feels like the right bet after seeing all 5 autopsies, why
- If it doesn't, what change would make it survive
```

## Tone

- Generative, not paralytic. The goal is mitigation, not retreat.
- Specific. "We didn't hire fast enough" → "We assumed we could hire a Series-B-experienced VP Eng in 90 days; that role takes 6–9 months."
- Honor the team's own pessimism. They've been carrying these worries silently — name them so they become addressable.