Skip to content

uxderrick/mece-skill

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

7 Commits
 
 
 
 
 
 

Repository files navigation

MECE Skill

Mutually Exclusive, Collectively Exhaustive — a validation and decomposition skill for AI agents.

Stop shipping specs with overlaps and gaps. MECE ensures every category, requirement, and plan item belongs to exactly one bucket and nothing is left off the list.

What it does

Mode Trigger Output
Validate Run /mece after creating any structured list Overlap report + gap report + corrected structure
Decompose /mece break down [topic] MECE issue tree from scratch

Why this matters

Most flawed thinking comes from two sources:

  1. Counting something twice under different names (overlap)
  2. Missing something entirely (gap)

This is expensive when you're writing product specs, designing APIs, planning roadmaps, or prompting AI to build features. MECE catches both before they become code, strategy, or shipped product.

Install

Quick install (recommended)

npx skills add https://github.com/uxderrick/mece-skill

Manual install

Clone the repo:

git clone https://github.com/uxderrick/mece-skill.git

Then copy to your agent's skills directory:

Agent Command
Claude Code mkdir -p ~/.claude/skills && cp -r mece-skill/skills/mece ~/.claude/skills/mece
Cursor mkdir -p ~/.cursor/skills && cp -r mece-skill/skills/mece ~/.cursor/skills/mece
VS Code / Copilot mkdir -p ~/.copilot/skills && cp -r mece-skill/skills/mece ~/.copilot/skills/mece
Codex (OpenAI) mkdir -p ~/.agents/skills && cp -r mece-skill/skills/mece ~/.agents/skills/mece
Gemini CLI mkdir -p ~/.gemini/skills && cp -r mece-skill/skills/mece ~/.gemini/skills/mece
Universal (all agents) mkdir -p ~/.agents/skills && cp -r mece-skill/skills/mece ~/.agents/skills/mece

Project-level install

Add to a specific repo instead of globally:

mkdir -p .agents/skills && cp -r mece-skill/skills/mece .agents/skills/mece

Usage

Validate existing work

Write your spec, feature list, or plan. Then:

/mece

The skill reviews your structure and reports:

  • Overlaps — items that belong to multiple categories
  • Gaps — scenarios not covered by any category
  • Suggested fix — corrected MECE structure

Decompose a new problem

/mece break down user authentication
/mece break down our pricing tiers
/mece break down the deployment pipeline

Validate inline (no slash command)

Just ask your AI agent:

"Check if this list is MECE" "Is anything overlapping or missing here?"

The skill auto-triggers on these patterns.

Who is this for?

  • Vibecoders — Stop generating code from specs with holes. Validate before you build.
  • Product managers — PRDs, user stories, and roadmaps that actually hold up in sprint planning.
  • Engineers — API designs, state management, and architecture with clean boundaries.
  • Founders — Strategy docs, pitch decks, and business plans with no logical gaps.
  • Consultants — You already know MECE. Now your AI does too.

Examples

See examples.md for full before/after walkthroughs:

  • The Protein Problem (everyday intuition)
  • The Packing List (collective exhaustiveness)
  • Product Feature Spec (PM workflow)
  • API Endpoint Design (engineering)
  • State Management / Reducers (React/frontend)
  • Roadmap Phases (product strategy)

Background

MECE was created by Barbara Minto at McKinsey & Company in the 1960s. It's been the foundational problem-structuring tool at McKinsey, Bain, and BCG for 60+ years.

This skill brings MECE to AI-native workflows — where the cost of overlaps and gaps is amplified because AI builds exactly what you tell it to, including the mistakes.

Inspired by this thread by @jayanaman.

License

MIT

About

MECE (Mutually Exclusive, Collectively Exhaustive) validation and decomposition skill for AI agents. Works with Claude Code, Cursor, Copilot, Codex, Gemini CLI.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors