← ClaudeAtlas

pre-mortemlisted

Runs a Klein pre-mortem on a project — assumes it already failed, brainstorms why, scores each failure mode by likelihood × impact, and attaches a mitigation plus an early-warning signal to the top risks. Use when launching a feature, kicking off a quarter, or before a big bet and you want to ask "imagine this shipped and flopped — what killed it?" or "where is this most likely to go wrong?"
Sidsaladi9/persona-os · ★ 5 · AI & Automation · score 81
Install: claude install-skill Sidsaladi9/persona-os
# Pre-Mortem A pre-mortem (Gary Klein) flips a postmortem forward in time: you assume the project has already failed, then work backward to explain why. This "prospective hindsight" frames failure as a given, which licenses honest, specific concerns that optimism normally suppresses — and turns them into mitigations before they cost you anything. **Grounded in:** *Performing a Project Premortem (HBR)* — Gary Klein: assume it failed, list why, rate likelihood × impact, mitigate the top modes. **Go deeper (The Product Channel):** [6 Fatal Mistakes to Avoid](https://sidsaladi.substack.com/p/week-42-6-fatal-mistakes-to-avoid) ## When to use this - You're about to commit to a launch, migration, or large bet and want to stress-test it before sinking resources. - The plan feels suspiciously smooth and nobody is raising objections in the room. - Kicking off a quarter or epic where the cost of a wrong assumption is high. - A senior stakeholder asked "what could go wrong?" and you want more than a hand-wavy risk slide. - You've seen similar projects fail before and want to name the pattern before it repeats. ## Before you start (gather these) - **The project + its definition of success**: what's shipping, by when, and the specific metric/outcome that means it worked. - **The plan or approach**: how you intend to get there (key milestones, dependencies, owners). - **Stakeholders and constraints**: who's involved, the deadline, budget, team capacity, technical or political constraint