Reading Time: 8 minutes

Hey Prompt Lover,

Let me ask you something.

When you take a big decision to Claude, maybe a pricing question, a positioning choice, whether to launch something or scrap it. What do you actually get back?

One answer.

One perspective. One angle. One set of reasoning that might be brilliant or might have a blind spot the size of a truck and you'd have no way to know because you only saw one view.

That's how most people use AI for decisions. And it's exactly why most AI-assisted decisions still feel like a coin flip dressed up in confident language.

What I'm about to give you fixes that.

It's called the LLM Council. It's one skill from the Claude Skills Ultimate Pack dropping Tuesday. And I'm giving it to you today free before the full pack even launches.

Because I want you to feel what one skill can do. Then I want you to imagine what the full pack brings.

Here's the idea behind it.

AI Agents Are Reading Your Docs. Are You Ready?

Last month, 48% of visitors to documentation sites across Mintlify were AI agents, not humans.

Claude Code, Cursor, and other coding agents are becoming the actual customers reading your docs. And they read everything.

This changes what good documentation means. Humans skim and forgive gaps. Agents methodically check every endpoint, read every guide, and compare you against alternatives with zero fatigue.

Your docs aren't just helping users anymore. They're your product's first interview with the machines deciding whether to recommend you.

That means: clear schema markup so agents can parse your content, real benchmarks instead of marketing fluff, open endpoints agents can actually test, and honest comparisons that emphasize strengths without hype.

Mintlify powers documentation for over 20,000 companies, reaching 100M+ people every year. We just raised a $45M Series B led by @a16z and @SalesforceVC to build the knowledge layer for the agent era.

Make Your Docs Agent-Ready

Andrej Karpathy — one of the most respected AI researchers alive figured out something that should have been obvious but wasn't. When you're facing a real decision with real stakes, one AI perspective isn't enough.

Different thinking styles catch different things. The answer you need isn't one confident response. It's the synthesis of multiple independent perspectives that have pressure-tested each other.

So he built a system where you dispatch a question to multiple AI models, have them peer-review each other anonymously, then a chairman produces a final verdict.

The LLM Council skill brings that entire methodology into Claude. One skill file. One trigger phrase. The whole process runs automatically.

Here's what happens when you use it.

You say "council this" — or "war room this" or "pressure-test this" or "I can't decide between X and Y" — and five independent advisors immediately go to work on your question. Each one thinks from a fundamentally different angle.

The Contrarian actively looks for what's wrong, what's missing, what will fail. Not a pessimist — the friend who saves you from a bad deal by asking the questions you're avoiding.

The First Principles Thinker ignores the surface question and asks what you're actually trying to solve. Sometimes the most valuable thing this advisor says is "you're asking the wrong question entirely."

The Expansionist looks for upside everyone else is missing. What could be bigger? What adjacent opportunity is hiding? What happens if this works even better than expected?

The Outsider has zero context about you, your field, or your history. They respond purely to what's in front of them. Experts develop blind spots. The Outsider catches the thing that's obvious to everyone else but invisible to you.

The Executor only cares about one thing — can this actually be done and what's the fastest path to doing it? Every idea sounds brilliant until the Executor asks what you do Monday morning.

These five advisors work simultaneously.

Independent.

No influence on each other.

Then they peer-review each other's responses anonymously.

Then a chairman reads everything and produces a final verdict.

Not "it depends." A real answer. Where the advisors agree. Where they clash. What they all missed. And the one concrete thing you should do first.

A real example so you can see this working.

Say you're deciding whether to launch a $297 course on Claude for beginners. Your audience is non-technical solopreneurs.

The Contrarian says the market is flooded, $297 competes with free YouTube content, and your non-technical audience means high support burden and refund risk.

The First Principles Thinker asks what you're actually trying to achieve — because if it's revenue, a course is one of the slowest paths, and if it's authority, a free resource might do more.

The Expansionist sees a massive underserved market. Everyone's teaching advanced stuff. Nail the beginner angle and you own the entry point to the entire space. $297 might actually be too low.

The Outsider says they don't know what Claude Code is. If they saw the course title they wouldn't know if it's for them. The name means nothing outside your world. Sell the outcome not the tool.

The Executor ignores all of it and says — before you build anything, run a live workshop at $97 to 50 people. Validate demand, get testimonials, create the raw material. If 50 people don't buy the workshop, 500 won't buy the course.

The chairman reads all five, synthesizes the peer reviews, and comes back with the real answer — don't build the course yet. Reframe entirely around the outcome not the tool. And the one thing to do first? Run a $97 live workshop called "How to automate your first business task with AI." Don't mention Claude Code in the title.

That's five angles. One clear direction. In one session.

How to use this.

How to install and use it. Right now. Today.

Step 1 — Download the skill file below. Save it to your computer.

Step 2 — In your Claude Code working folder, create this path if it doesn't exist already: .claude/skills/llm-council/

Step 3 — Drop the downloaded SKILL.md file into that folder.

Step 4 — Open Claude Code in your project folder and type "council this" followed by any real decision you're sitting on.

That's it. The skill activates automatically on the trigger phrases. It reads your existing context files so the advisors already know your business, your audience, and your constraints before they say a word.

Every council session produces two files automatically — a clean HTML report with the verdict, where advisors agreed and clashed, and the one concrete next step. And a full transcript if you want to go deeper.

Now here's what I need you to understand.

This is one skill.

Tuesday's pack has Cowork skills. Claude Code skills. Video skills. Finance and day trading skills. And more that I haven't announced yet.

If this one skill changes how you make decisions and it will. I want you to sit with that for a second before Tuesday.

Because this is a taste. Not the meal.

Install it today. Take your hardest current decision to the council. See what five perspectives and a chairman give you that one answer never did.

Then reply and tell me what it surfaced.

If there's a skill you want added to Tuesday's pack, reply now. The finance skill made it in from one email. The pack isn't completely locked yet.

Tuesday. Full drop. Stay close.

— Prompt Guy