How do I install the Pmstack Premortem skill?

Run npx skillmds@latest add ryanalberts/pmstack-premortem in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Pmstack Premortem safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Pmstack Premortem?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Pmstack Premortem free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Pmstack Premortem?

RyanAlberts (@ryanalberts) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Pmstack Premortem by ryanalberts — Agent Skill

Name: Pmstack Premortem
Author: RyanAlberts

Pmstack Premortem

Run a Klein-style pre-mortem on a draft PRD before launch — simulates 3 plausible failure stories 6 months in the future, names leading indicators and mitigations, then offers to mutate the PRD's Risks section. Trigger when the user says "pre-mortem," "what could go wrong," "risk pass," "imagine this fails," "stress test the spec," or asks to harden a PRD before review. Use after /prd, before launch.

RyanAlberts 4fbac59 2.5 KB Updated Sep 17, 2026

File contents

premortem — pre-mortem on a target PRD

Klein's pre-mortem trick: pretend the feature failed 6 months from now and write the post-mortem now, before launch. The only pmstack capability that mutates an existing artifact rather than appending another.

Required structure

For each of 3 failure stories:

### Failure story <N>: <one-line headline>
**What happened:** <2–4 sentences narrating the specific failure>
**Anchored to PRD:** <quote or section reference from the source PRD>
**Leading indicator:** <a metric, signal, or qualitative observation that would have surfaced this BEFORE launch>
**Mitigation:** <one specific action to take before shipping>

The 3 stories must span at least: (a) capability/quality, (b) adoption/UX, (c) operational/cost. Do not stack 3 quality failures.

Hard rules

Stories must be specific, not generic. "Adoption could be lower than expected" fails. "Devs share screenshots of bad code-review suggestions on day 1, the meme spreads, the trial closes sour" passes.
Each leading indicator must be measurable today. No reliance on production traffic that doesn't exist — use a pre-launch proxy.
Confirmation gate is mandatory. After writing the artifact, list the 3 mitigations and ask which to append to the PRD's Risks section. Do not auto-mutate. Wait for explicit selection ("1,3" or "all" or "none").
Anchor every story to the PRD. Quote or section reference required.

Where to write

With filesystem: outputs/premortem-<topic-slug>-<YYYY-MM-DD>.md. On accept, mutate the source PRD's Risks section. Rejected mitigations land in ## Rejected mitigations of the premortem artifact.
Inline (web/mobile): emit the premortem markdown plus a separate "Suggested PRD diff" block the user pastes into their PRD.

Decision-log entry

- <date> — premortem(<topic>): <N>/3 risks accepted — outputs/premortem-<topic>-<date>.md

Tone

Concrete, narrative, blunt about failure modes. PMs systematically under-fill risk sections; this skill is the corrective.

Cold-start behavior

Works the moment one PRD exists. No history required.

---name: pmstack-premortemdescription: Run a Klein-style pre-mortem on a draft PRD before launch — simulates 3 plausible failure stories 6 months in the future, names leading indicators and mitigations, then offers to mutate the PRD's Risks section. Trigger when the user says "pre-mortem," "what could go wrong," "risk pass," "imagine this fails," "stress test the spec," or asks to harden a PRD before review. Use after /prd, before launch.---# premortem — pre-mortem on a target PRDKlein's pre-mortem trick: pretend the feature failed 6 months from now and write the post-mortem now, before launch. The only pmstack capability that **mutates** an existing artifact rather than appending another.## Required structureFor each of 3 failure stories:```### Failure story <N>: <one-line headline>**What happened:** <2–4 sentences narrating the specific failure>**Anchored to PRD:** <quote or section reference from the source PRD>**Leading indicator:** <a metric, signal, or qualitative observation that would have surfaced this BEFORE launch>**Mitigation:** <one specific action to take before shipping>```The 3 stories must span at least: (a) capability/quality, (b) adoption/UX, (c) operational/cost. Do not stack 3 quality failures.## Hard rules- **Stories must be specific, not generic.** "Adoption could be lower than expected" fails. "Devs share screenshots of bad code-review suggestions on day 1, the meme spreads, the trial closes sour" passes.- **Each leading indicator must be measurable today.** No reliance on production traffic that doesn't exist — use a pre-launch proxy.- **Confirmation gate is mandatory.** After writing the artifact, list the 3 mitigations and ask which to append to the PRD's Risks section. Do not auto-mutate. Wait for explicit selection ("1,3" or "all" or "none").- **Anchor every story to the PRD.** Quote or section reference required.## Where to write- With filesystem: `outputs/premortem-<topic-slug>-<YYYY-MM-DD>.md`. On accept, mutate the source PRD's Risks section. Rejected mitigations land in `## Rejected mitigations` of the premortem artifact.- Inline (web/mobile): emit the premortem markdown plus a separate "Suggested PRD diff" block the user pastes into their PRD.## Decision-log entry`- <date> — premortem(<topic>): <N>/3 risks accepted — outputs/premortem-<topic>-<date>.md`## ToneConcrete, narrative, blunt about failure modes. PMs systematically under-fill risk sections; this skill is the corrective.## Cold-start behaviorWorks the moment **one** PRD exists. No history required.

Frequently asked questions