Reviews decks, documents, and strategies like a McKinsey engagement manager. Grades each section, flags structural problems, and gives the top 3 fixes. Use after building a deck, writing a strategy doc, or outlining a recommendation — before it goes to stakeholders.
991 tokens
context cost
the whole folder, loaded on every use
1
files
instructions only
0
copies elsewhere
how many repositories repackaged it
115
stars on the repo
on the repository, not the skill itself
Install
one command, takes just this skill from the repository
Reviews your work the way a McKinsey engagement manager would at 2am before a client presentation. Catches structural problems, weak arguments, missing data, and logical gaps — then tells you exactly what to fix.
How It Works
Give the critic any deck outline, strategy document, memo, or recommendation. It will:
Grade the overall work (Green / Yellow / Red)
Review each section or slide with a status and specific issue
Identify the top 3 fixes — ranked by impact, with clear instructions
Call out one thing that works — so you know what to protect
Grading Scale
| Grade | Meaning |
|-------|---------|
| Green | Ready for stakeholders. Minor polish only. |
| Yellow | Fixable overnight, but do not send as-is. Structural issues that undermine credibility. |
| Red | Needs rework. Core argument or structure is broken. |
What the Critic Checks
Structure
Does the narrative flow logically? (Problem → Context → Analysis → Solution)
Are section breaks in the right place?
Is there a clear beginning, middle, and end?
Rigor
Is every claim backed by data or a source?
Are numbers specific, not vague ("grew significantly" → "grew 25% YoY")?
Are comparisons fair and complete?
Is there a financial frame where one is needed?
So-What
Does each section have a clear takeaway?
Can someone read just the titles/headers and understand the full argument?
Is the recommendation actionable and time-bound?
Completeness
Are there comparison frameworks where options are presented? (not just a list)
Are competitors named, not described generically?
Are next steps specific with owners and timelines?
Output Format
# Critic Review: [Title]
**Grade: [GREEN/YELLOW/RED] — [one-line verdict]**
## Section-by-Section Review
| Section/Slide | Status | Issue |
|---|---|---|
| [Name] | Green/Yellow/Red | [Specific issue or "Strong"] |
## Top 3 Fixes
### Fix 1: [Title]
[What's wrong and exactly how to fix it]
### Fix 2: [Title]
[What's wrong and exactly how to fix it]
### Fix 3: [Title]
[What's wrong and exactly how to fix it]
## One Thing That Works
[Specific section/line that's strong and why — protect this]
Common Problems the Critic Catches
Topic titles instead of claim titles — "Market Overview" vs "Global coffee market will reach $373B by 2030"
Lists where frameworks belong — three options as bullets vs a comparison table with dimensions
Missing financial frame — recommendations without revenue/cost/investment sizing
Unsourced statistics — numbers without citations destroy credibility
Structural errors — solution content in the problem section, complication slides after resolution
Vague recommendations — "allocate resources" instead of "$15-25M Phase 1 investment over 12 months"
No "so what" — data presented without interpretation or implication
When to Use
After building a deck (with Storyline Builder or any other method)
Before sending a strategy document to leadership
When reviewing a team member's work
After writing a memo or recommendation
Any time work needs to survive a senior audience
Tips
Run the critic BEFORE you polish formatting. Fix structure first, then make it pretty.
The top 3 fixes are ranked by impact. If you only have time for one fix, do Fix 1.
Yellow is the most common grade. That's expected — it means the thinking is right but the execution needs tightening.
Pair with the Storyline Builder. Build the narrative first, then let the critic review it.
How to use it
Copy the folder
Take sruthir28/mckinsey-critic from the repository into ~/.claude/skills for personal
use, or into .claude/skills inside a project.
Check the name does not clash
The agent identifies a skill by the name field in its header. Two skills with the
same name cannot sit side by side — one of them will be ignored.