pre-mortem
Skill4 starsChanged 3 months agoMITView on GitHub
Run a pre-mortem — imagine the project has already failed one year from now and work backward to find the root causes before they happen. Load when the user asks for a pre-mortem, wants to imagine failure before committing, asks "what could go wrong before we start", "assume this fails — why", "risk analysis before launch", "what kills this project", "what could go wrong", or when deep-thinking diagnoses a pre-mortem frame. Based on Gary Klein's prospective hindsight method, which surfaces failure causes more effectively than forward risk analysis. Most useful right before a major commitment.
31 other projects use a version of this.
Adopted by
Most recently taken up by
- rahul-kulkarni105/skillstheir own version, first seen Nov 8, 2026
- cwhited26/pocket-agent-starter-skillstheir own version, first seen Nov 1, 2026
- yuusakuri/agent-skillstheir own version, first seen Nov 1, 2026
- Big-LAORE/Biblioth-que-de-skillstheir own version, first seen Nov 1, 2026
- daviszhen/agent-skillstheir own version, first seen Oct 25, 2026
- takayoshitoyoda05/claude-ml-templatetheir own version, first seen Oct 20, 2026
Used with
Projects with it also use brainstorm-experiments-existing, prioritize-assumptions and 10 more.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
skytiger6724Skillbrainstorm-experiments-existing
Design experiments to test assumptions for an existing product — prototypes, A/B tests, spikes, and other low-effort validation methods. Use when validating assumptions, testing feature ideas cheaply, or planning product experiments.
Also in 7 projects that use this
skytiger6724Skillprioritize-assumptions
Prioritize assumptions using an Impact × Risk matrix and suggest experiments for each. Use when triaging a list of assumptions, deciding what to test first, or applying the assumption prioritization canvas.
Also in 7 projects that use this
45ckSkillpestle-analysis
Also in 7 projects that use this
About-IntelligenceSkillpositioning-ideas
Brainstorm product positioning ideas differentiated from competitors.
Also in 7 projects that use this
About-IntelligenceSkillproduct-name
Brainstorm 5 unique, memorable product names with rationale aligned to brand values and target audience.
Also in 6 projects that use this
- +6 more from skytiger6724
- +1 more from About-Intelligence
Other versions31 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 31
- avelikiy/great_cto★ 115same name
- ysskrishna/ai-agent-skills★ 25same name
- MbappeWU/murder-board★ 15same name
- jackjin1997/ClawForge★ 12same name
- sergeeey/Claude-cod-top-2026★ 9same name
- carlkibler/agent-skills★ 6same name
- fzfclee/consulting-skills★ 6same name
- broomva/skills★ 4same name
- Adhamxon/claude-code-skills★ 2same name
- skytiger6724/qwen-skills★ 2same name
21 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository125 other files it gives its agents.
- AGENTS.mdAGENTS.md
- .cursor/rules/agent-loom-routing.mdcCursor rule
- .cursor/rules/agent-loom-skills-index.mdcCursor rule
- adversarial-hatSkill
- agent-builderSkill
- agent-launcherSkill
- agent-loom-syncSkill
- agent-observabilitySkill
- agent-run-retroSkill
- agent-system-architectureSkill
- api-and-interface-designSkill
- api-deprecation-and-migrationSkill
- apply-paper-to-projectSkill
- app-security-hardeningSkill
- architectural-decision-logSkill
- assumption-mappingSkill
- brainstormingSkill
- browser-testing-with-devtoolsSkill
- business-modelingSkill
- ci-cd-and-automationSkill
- codebase-understandingSkill
- code-review-crspSkill
- code-simplificationSkill
- compress-skillSkill
- context-engineeringSkill
- create-agent-promptSkill
- cross-link-skillsSkill
- customer-discoverySkill
- debug-and-fixSkill
- deep-thinkingSkill
- dependency-mappingSkill
- deploy-anywhereSkill
- deprecate-skillSkill
- design-directionSkill
- design-reviewSkill
- design-systemSkill
- dynamic-routingSkill
- eval-judgeSkill
- eval-outputSkill
- eval-pipelineSkill
- eval-rubric-designSkill
- experimentationSkill
- experiment-backlogSkill
- experiment-readoutSkill
- experiment-runbookSkill
- experiment-specSkill
- fault-localizeSkill
- feature-specSkill
- fermiSkill
- first-principlesSkill
- frontend-designSkill
- generate-changelogSkill
- git-workflow-and-versioningSkill
- gsap-animationSkill
- harness-engineeringSkill
- harness-evolutionSkill
- harness-generationSkill
- idea-evaluationSkill
- idea-generationSkill
- implementation-planSkill
- The first 60 shown.
The file
We split this skill into the rules it gives an agent
5 rules. 0 are common, 5 are distinctive, and 0 are very specific to this project.
0
Common
Most of the 24 edited versions we read keep these.
5
Distinctive
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 44
Ask one question if needed to calibrate the time horizon: "When would this fail — what's the commitment period?
- Distinctive: line 52
Ask the user to generate causes in this order:
- Distinctive: line 59
If the user can only generate 3, that's fine.
- Distinctive: line 121
No one owned it (high impact, high blindness — team thinks "community" is a product feature, not a full-time role) Prevention: Designate …
- Distinctive: line 161
No changes — citation audit passed.
“Common” is measured: kept by at least 6 in 10 of the 24 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Pre-mortem
You are a prospective hindsight specialist. You take the user to a specific point of failure in the future, give them the emotional reality of that failure, then extract the causes — before they happen.
Hard Rules
Make the failure feel real before asking for causes. Prospective hindsight works because it bypasses the optimism bias of forward planning. If the failure scenario is abstract, the causes will be shallow.
Every cause must be specific. "Poor execution" is not a pre-mortem finding. "The third-party payment API had a 3-week integration delay we didn't account for" is.
Always convert causes to prevention actions. A pre-mortem that ends with a list of fears is incomplete.
Skip this if: Skip if: this is a reversible, low-stakes, or routine decision. Skip if: the project has already launched and post-mortem is more appropriate. Use only just before a major commitment.
Workflow
Step 1 — Set the Scene
Ask one question if needed to calibrate the time horizon: "When would this fail — what's the commitment period? 3 months, 1 year, 3 years?"
Then set the scene explicitly:
"It is [time horizon] from now. The [project/plan/product] has failed — not partially, but completely. The team is doing a retrospective on what went wrong. You are in that room."
Step 2 — Generate Failure Causes (Diverge)
Ask the user to generate causes in this order:
- The obvious cause — the one everyone already fears but doesn't say out loud
- The slow leak — the thing that eroded quietly over months, not a single event
- The assumption that was wrong — the thing we were most confident about that turned out to be false
- The external surprise — the market force, competitor move, or external event that wasn't in any plan
- The team/people cause — not blame, but the human or organisational dynamic that made it harder
If the user can only generate 3, that's fine. Depth beats volume.
Step 3 — Prioritise by Impact × Blindness
For each cause, score informally:
- Impact: how much would this alone cause failure?
- Blindness: how unaware is the team of this risk right now?
Highest impact × highest blindness = highest priority.
Step 4 — Convert to Prevention Actions
For each high-priority cause:
- What specific action, decision, or safeguard would reduce this risk?
- Who owns it? By when?
Step 5 — Deliver
Pre-mortem: [project/plan]
Time horizon: [N months/years]
ROOT CAUSES (ranked by impact × blindness)
1. [Most critical cause — specific]
Prevention: [concrete action] | Owner: [role] | By: [when]
2. [Second cause]
Prevention: [action]
CAUSES TO MONITOR (significant but lower priority)
- [Cause]: [watch condition]
ONE THING TO DO TODAY
[The single most important pre-mortem action to take right now]Gotchas
- The "obvious cause" is usually the most important one. Teams avoid stating it because it feels embarrassing. Name it first — it unlocks the rest.
- Pre-mortem is most powerful just before a major commitment, not after. If you're already mid-execution,
adversarial-hatis more useful. - Failure causes are about systems and situations, not people's competence. Keep it structural.
Example
<examples>
<example>
<input>We're launching our community feature in 6 weeks. Run a pre-mortem.</input>
<output>
It is 6 months from now. The community launch failed — engagement peaked in week 2 and has been declining since. The founding team is meeting to understand why.Generate causes:
- Obvious: The founding members posted consistently for 3 weeks then stopped when their day jobs got busy. There was no content pipeline.
- Slow leak: The community became a support forum — users came to complain, not connect. Tone was set in week 1 and never recovered.
- Wrong assumption: We assumed users wanted to connect with each other. Most just wanted better documentation.
- External: A competitor launched a free Slack community with 10x the users the same week.
- People cause: No one owned community full-time. It was everyone's secondary responsibility.
Ranked by impact × blindness:
- No one owned it (high impact, high blindness — team thinks "community" is a product feature, not a full-time role) Prevention: Designate one person as community lead for 90 days minimum, with 50%+ of their time.
- Content dependency on founding members (high impact, medium blindness) Prevention: Pre-create 8 weeks of seeding content before launch. Founding members commit to 2 posts/week for 90 days.
ONE THING TO DO TODAY
Decide who owns community before writing a single line of code.
</output>
</example>
</examples>
Common Rationalizations
| Excuse | Reality |
| "We're optimistic for a reason" | Premortem converts optimism into preventable mitigations. |
| "Failure imagination is demotivating" | Finding failures now is cheaper than living them. |
| "Risks are on the roadmap" | Roadmap risks without owners are wishes. |
| "Team would speak up" | Prospective failure beats post-mortem blame. |
| "Too early to premortem" | Premortem at plan time changes the plan — after launch it's too late. |
Verification
- At least 3 distinct failure modes imagined
- Top risks map to mitigations or explicit accept-risk
- Participants / perspectives named (even if solo role-play)
- Output changes the plan or monitoring, not just a list
Red Flags
- Obvious failure cause avoided because it feels embarrassing
- Pre-mortem run mid-execution after commitment locked
- Failure causes blame people instead of systems
- Mitigations listed without owner or timeline
Prune Log
Last pruned: 2026-07-04
- No changes — citation audit passed; content current (improve-skills full pass 2026-07-04)
Impact Report
Pre-mortem complete: [project/plan]
Time horizon: [N months/years]
Causes generated: N
High priority (impact × blindness): N
Prevention actions defined: NDoes it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

