Guide · Agent skill
How to use the Pre-Mortem Risk Analysis agent skill
Updated · written with AI from the project's own documentation and checked against our catalogue data
The pre-mortem agent skill, published as phuryn/pm-skills under pm-execution/skills/pre-mortem, asks an AI agent to review a product requirements document or launch plan by assuming the launch has already failed. It is written for product managers, founders and small teams who want a structured risk review before committing resources. After this guide you will know what the skill does, how to install it, how to ask for it in plain language, and what to check before trusting the output.
Pre-Mortem Risk Analysisphuryn/pm-skills · MIT Download the skill Quick start
- 1Download the ZIP. Get the pre-mortem skill ZIP from this page. It contains SKILL.md.
- 2Unzip into a skills folder. Place the pre-mortem folder in ~/.claude/skills/ for personal use, or in a project's .claude/skills/ folder.
- 3Or upload in the Claude apps. In the Claude apps, go to Settings → Capabilities → Skills and upload the skill.
- 4Attach your PRD or launch plan. Share the document or paste its product description, target market, assumptions and timeline.
- 5Ask for a pre-mortem. Use a plain request such as asking for a pre-mortem or a stress test of your launch plan.
- 6Review Tigers, Paper Tigers and Elephants. Check each category, then go through the launch-blocking action plans with their owners and dates.
- 7Revisit before launch. Run the review again two to three weeks before launch to confirm mitigations are on track.
What the pre-mortem skill makes the agent do
When you ask the agent to review a PRD or launch plan, the skill tells it to imagine the product launches in 14 days and then fails: customers don't adopt it, revenue targets miss, or reputation takes a hit. The agent works backward from that failure to find what went wrong, what was missed, and what the team was overconfident about.
The agent then sorts each possible problem into one of three groups. Tigers are real problems that could derail the project. Paper Tigers are concerns that sound valid but are overblown. Elephants are unspoken worries that nobody on the team is validating and that deserve investigation.
- Tigers: based on evidence, past experience or clear logic; require action.
- Paper Tigers: worth documenting to align stakeholders, but not worth significant resources.
- Elephants: could be real; the agent recommends an investigation approach.
- Tigers are also classified as launch-blocking (solve before launch), fast-follow (solve within 30 days after launch) or track (monitor and solve if it becomes an issue).
Install the skill from the ZIP download
Download the ZIP for this skill from this page. The skill package contains one file, SKILL.md. Unzip it so that the pre-mortem folder, containing SKILL.md, sits inside your skills directory. For a personal install, use ~/.claude/skills/. For a single project, use the .claude/skills/ folder inside that project.
If you use the Claude apps instead of files on disk, you can upload the skill under Settings → Capabilities → Skills.
- Personal install: place the pre-mortem folder in ~/.claude/skills/.
- Project install: place it in the .claude/skills/ folder of your project.
- Claude apps: upload the skill under Settings → Capabilities → Skills.
Trigger the skill with natural requests
You do not need a special command. The skill's description tells the agent when to use it: preparing for launch, stress-testing a product plan, or identifying what could go wrong. Ask in those terms and attach or paste your document.
The skill's instructions say to read the PRD if you provide a file, so pointing the agent at the document works best. The skill also says the output should be saved as a markdown file named PreMortem-[product-name]-[date].md.
- "Run a pre-mortem on this PRD before we launch the new onboarding flow."
- "Stress-test our launch plan and tell me what could go wrong."
- "Review this product plan for hidden risks engineering and design may not be raising."
- "Sort the risks in this launch doc into real threats, overblown worries and unspoken concerns."
What to provide and what the output looks like
The skill lists no required inputs. In practice the quality of the result depends on the material you give it. Include the product description, target market, key assumptions and timeline from your PRD or launch plan, because the skill's instructions tell the agent to focus on these.
The skill specifies a structure for the output: Tigers (Real Risks), Paper Tigers (Overblown Concerns), Elephants (Unspoken Worries), and Action Plans for Launch-Blocking Tigers. Each action plan lists the risk, mitigation, owner and due date. The skill also says to revisit the pre-mortem two to three weeks before launch to check that mitigations are on track.
Safety notes: scripts, network and sources
This skill runs no scripts and uses no network access, according to its listing. The skill's instructions mention web search for competitive or market research, but only if the agent decides it is relevant. The skill does not name any specific sources, so the agent's findings are not verified evidence.
The output is based on the agent's reasoning and whatever you supply. Treat it as a structured discussion aid, not as a fact check. Competitor claims and market assumptions need checking by you or your team.
- Runs scripts: no.
- Uses network: no (web search is used only if the agent judges it relevant).
- Requires: nothing listed.
- Not a substitute for real market research or customer validation.
Tips for a useful pre-mortem
Give the agent cross-functional context. The skill's notes ask it to involve engineering, design and go-to-market perspectives, so include what those teams are worried about if you know it. The skill also says to default to Tiger when unsure, so expect a cautious list and review it rather than accepting everything.
Keep the output file. The saved markdown is the artifact you can share and revisit before launch.
- Paste the real assumptions your plan depends on; vague input gives vague risks.
- Review every Paper Tiger, since the agent's dismissals are its reasoning, not proof.
- Assign real owners to the launch-blocking action plans before you leave the meeting.
- Rerun the review two to three weeks before launch to check progress.
What people use it for
- A product manager stress-testing a feature PRD before an engineering commitment.
- A founder checking a go-to-market plan for risks the team has not discussed.
- A cross-functional team collecting engineering, design and go-to-market concerns in one saved document.
- Preparing a list of launch-blocking items with owners and dates before a launch review meeting.
Frequently asked questions
Is the pre-mortem agent skill free?
The source lists the skill under the MIT licence, which permits use under its terms. The source does not state pricing for the Claude apps or any other platform, so check those separately.
What do I need to run the pre-mortem skill?
The skill's listing says nothing is required. You need an agent that can load skills, such as the Claude apps or a setup that reads .claude/skills folders, plus your PRD or launch plan.
Where do I install the pre-mortem skill?
Unzip the downloaded package so the pre-mortem folder sits in ~/.claude/skills/ for personal use or in a project's .claude/skills/ folder. In the Claude apps, upload it under Settings → Capabilities → Skills.
Does the skill check facts or competitor claims for me?
No. The output is based on the agent's reasoning and what you supply. Web search is used only if the agent decides it is relevant, and the skill does not specify sources, so verify competitor and market claims yourself.
Can the pre-mortem replace market research or customer validation?
No. The skill is a structured risk discussion aid. Elephants in particular call for real investigation, such as validating assumptions with customers or data.
What is the difference between a Tiger and a Paper Tiger?
A Tiger is a real problem based on evidence, experience or clear logic that requires action. A Paper Tiger is a concern that seems valid on the surface but is unlikely or overblown, so it is documented rather than heavily resourced.
New agent skills, repos and AI tools — every week
The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.
Our verdict
Use this skill if you are a product manager, founder or small team who wants a structured, cross-functional risk review of a PRD or launch plan before committing resources. Pick something else if you need verified market research, customer validation or fact-checked competitor analysis, since the skill produces reasoning based on what you supply.
Download the skill