You know that thing where you ask one AI one question, get one confident answer, and have absolutely no way to tell if it is brilliant or quietly wrong? That is the trap. A single AI gives you a single take, often delivered with total confidence even when it is missing something obvious. And for the decisions that actually matter (your pricing, your positioning, whether to take the deal), one confident answer is not enough.
This fixes that. It is a free skill called the LLM Council, built by Ole Lehmann and adapted from Andrej Karpathy's "LLM Council" methodology. I call it the Claude Council in my reel because it runs entirely inside Claude, but its real name is the LLM Council, so do not get confused when you go to install it. Instead of one answer, it runs your question through 5 independent advisors, has them anonymously review each other's answers, then a chairman pulls it all into one clear verdict. It is like having a board of advisors on call, for free, any time you are stuck.
how it worksHere is the whole thing in plain language. Four steps, all happening inside one Claude chat.
These are thinking lenses, not job titles. Each one is deliberately one-sided so that together they cover your blind spots.
The chairman's verdict comes back in five clean parts, so you can read it like a map:
It is completely free, and it is beginner friendly. You do not need to know any code. Skills work on every Claude plan, Free through Enterprise. Pick whichever path feels easier.
👉 Prefer to grab the file and upload it yourself? Download llm-council-skill.zip, then in the Claude app go to Customize, then Skills, then Add, then Upload a skill, and drag the file in.
The easy path: open a new chat in Claude and paste this exactly.
Please install this Claude skill for me. The SKILL.md file lives in this GitHub repo: https://github.com/aiwithremy/claude-skills-llm-councilClaude reads the repo and sets it up for you. Done.
The upload path: go to the repo, download the file called SKILL.md, then tell Claude:
I just downloaded a file called SKILL.md for the LLM Council skill, can you install it for me?In the Claude web or desktop app you can also go to Customize, then Skills, hit Add (next to Browse), choose Upload a skill, and drag the file in. The repo lives here if you want to look first: https://github.com/aiwithremy/claude-skills-llm-council
Once it is installed, you summon it with any of these trigger phrases. Just type one and add your question.
ask the council
council this: [your question]
run the council on [your question]
pressure-test this
stress-test this
war room this
debate thisThis is the part most people skip, and it is the difference between generic advice and a real war room.
This is your copy-and-paste library. Find the situation closest to yours, grab the prompt, swap in your real numbers, and run it. The more honest detail you add, the better the verdict.
Picking your one lead platform when you are spread thin.
council this: I'm a solo creator posting AI content. Right now I split my time across Instagram reels, LinkedIn carousels, and Threads, and I'm doing all three badly because I'm exhausted. My goal this year is to build an audience I can eventually sell a $47/mo community to. Should I go all in on one platform for the next 90 days, or keep a lighter presence on all three? If one, which one, given the community-sale goal?Repositioning when your hook audience is not your buyer audience.
council this: My most viral reels are fear-based hooks that pull huge reach but mostly attract scared employees who never buy anything. The people who actually message me about working together are small business owners who want AI to run their ads. Do I lean harder into the fear hooks for reach and accept low conversion, pivot my whole positioning toward business owners and lose the easy reach, or try to run both and risk a confused brand?Whether to take a sponsorship that does not fit your niche.
council this: A meal-kit brand offered me $1,400 for two reels. I make AI and marketing content, so meal kits are off-topic. The money is real, but my last three posts have been tightly on-niche and my engagement is finally climbing. Do I take the deal and risk diluting the audience I just tightened, negotiate for an AI angle they probably won't accept, or turn it down? This is only my second paid offer ever.Pricing a sponsored reel package without a benchmark.
council this: A SaaS brand wants a 3-reel package plus 30 days of usage rights to run my videos as their own ads. My only prior paid deal was $1,900 for two reels with no ad usage. They opened by asking for "your package rate." I'm tempted to quote $3,500 but I'm scared of pricing myself out of my first package deal. What number do I send, how do I structure the usage rights, and how do I justify it in the email?Killing or doubling down on a content series.
council this: I'm 8 episodes into a numbered series. Episodes 1 to 3 did great, then 4 through 8 flattened to half the views. My rule was to re-evaluate at 10 episodes. Do I push to 10 as planned and then decide, kill it now and free up the slots for fresh ideas, or refresh the format and treat the next two as a test?Choosing your call-to-action mix across reels.
council this: Most of my reels end with a hard CTA: "comment KEYWORD and I'll DM you the resource." My reach has dipped and I'm wondering if the constant keyword-bait comments are training the algorithm to see my engagement as low-quality. Do I keep the hard CTA because lead capture is the point, cut back to half and let the rest be pure value to protect reach, or test a soft "follow for more" close on my highest-reach posts only?A tripwire price point that is killing conversions.
council this: I sell an AI prompt pack as a low-ticket tripwire at $27 and conversion from my email list is about 2%. Some people in my niche price these at $9 to $17 and claim higher take rates. Do I drop to $17 to get more buyers into my world so I can upsell my $497 offer later, or hold $27 because the higher price filters for serious buyers? Am I even thinking about this the right way?Productize a service into a flat-price offer.
council this: I keep getting DMs asking me to set up their AI content system, and I quote it custom every time, usually $400 to $900. The scoping conversations are exhausting. I'm considering one fixed productized offer at a flat $397, same deliverables every time. The risk is some setups are more work than others and I'll lose money on the hard ones. Should I flatten to one price, do tiers, or keep quoting custom?Launch model versus evergreen.
council this: I'm deciding how to sell my paid community. Option A is a launch model: doors open for one week a few times a year with urgency. Option B is evergreen: doors always open, sell continuously with no urgency. I post daily so I have steady reach, but I hate the dead months a launch model creates, and I worry evergreen signups will trickle and never spike. Which fits a daily-posting creator with my reach?Free tier versus paid-only entry.
council this: I'm launching a community for marketers learning AI. Option A: open a free tier for social proof and nurture, then upsell paid later. Option B: skip free and go paid-only at a low launch price so the member count is all real buyers. Free risks a ghost-town of non-buyers; paid-only risks a tiny visible count at launch that kills momentum. Which structure should I open with?Anchor a high-ticket offer you have never sold.
council this: I want to add a 1:1 AI strategy call as my highest-priced offer, capped at 4 a month. I've never sold a call at any price. Part of me wants to anchor at $750 to position it as premium, part of me wants to start at $297 to get the first bookings and build testimonials. The calls also become content and case-study fuel, which matters more to me right now than the cash. Where should I set the price, and is launching this even the right move before I have proof?Choosing a positioning angle when you keep flip-flopping.
council this: I run a small Meta ads service for local gyms. Option A: "I get fitness studios fully booked with paid ads" (outcome promise, scary to back up because results vary). Option B: "I test 20 ad variations a week so your best ad is always live" (process promise, defensible but less sexy). My fear is the process angle sounds boring to gym owners who only care about more members. Which angle do I lead with on my site and in DMs?Reviewing a landing page hero before you run traffic.
council this, here is the hero section of my landing page and I want you to find what's weak before I run traffic. Headline: "Stop drowning in busywork. Let AI run the boring parts of your business." Subhead: "I build custom automations that save founders 10+ hours a week." Button: "Book a free audit." Context: cold traffic from my reels, audience is non-technical solopreneurs, the page gets visits but almost nobody clicks the button. What's killing the conversion?Stress-testing an Instagram bio that is not converting.
council this, here's my Instagram bio and I think it's why visitors aren't following or clicking my link. Bio: "AI + Marketing | Helping you work smarter not harder | Free AI toolkit below." Context: I get 5 to 15k views on reels but my follow rate from profile visits is under 2% and almost nobody taps the link. Is the bio the problem, and if so what's the weak part: the positioning, the "work smarter not harder" cliche, or the offer?Choosing the hook for a sales email subject line.
council this: I'm sending the launch email for a $497 two-week AI sprint to 600 people who downloaded my free toolkit. Three subject lines, I can send one. A: "Your first AI automation, built with you in 2 weeks." B: "Most people will never get AI to actually work for them. Here's the shortcut." C: "Open if you're tired of saving AI posts and never building anything." They barely know me. Which one, and what's the risk with C, which I'm leaning toward?Quit the day job to go full-time on the side hustle.
council this: I have a 9 to 5 paying $95k that I'm bored in, and a side business making $2,800/month growing about 15% month over month. I have about 4 months of expenses saved. Part of me wants to quit now to go all in, part says wait until it matches my salary. Should I quit now, set a revenue milestone to quit at, or negotiate going part-time first?Hire your first person versus stay solo.
council this: My business makes about $9k/month solo and I'm maxed on hours, turning down work. I could hire a part-time contractor at roughly $1,800/month to take admin and delivery off me, but it would eat most of my profit for a few months and I've never managed anyone. Do I hire now and bet on growth, raise my prices first to thin the workload, or push through solo for another quarter?Where to put your limited time and energy.
council this: I have about 15 free hours a week outside my job and I keep splitting them three ways: building my own business, studying for a certification that could get me promoted, and freelancing for quick cash. I'm making slow progress on all three and feeling burnt out. If I concentrated those 15 hours on the one path with the best 12-month payoff, which should it be, assuming I want long-term independence?Walk away from a draining client or keep the money.
council this: One client is about 45% of my freelance income but they're disrespectful, scope-creep constantly, and dealing with them is wrecking my motivation. Replacing that income would take a few months of pipeline-building I haven't done. Do I fire them now and eat the gap, set firm boundaries and a rate increase and see if they leave, or keep them while I quietly build replacement income first?Hire versus automate the bottleneck.
council this: My biggest bottleneck is the structural cut of my videos, which eats hours every week. Option A: hire an assistant at around $300 to $500 a month to take it off me. Option B: keep doing it myself because the cut is a taste call and I worry handing it off will make my reels feel generic. What's the right move, and if it's hiring, how do I hand off a taste-heavy task without losing my voice?When to kill a stalled side project.
council this: I built two small automation tools months ago that are now stale and sending out-of-date content. I keep saying I'll fix them when I have capacity, but that capacity never comes because filming and editing always win. Do I fully kill them, do a one-time rebuild, or leave them dormant and revisit next quarter? My real constraint is that content production is my number one lever right now.Picking the production tool stack.
council this: I'm standardizing my always-on automation layer and I keep going back and forth. My builds are mostly marketing workflows: scheduled content jobs, scraping AI news, batch generation, lead routing. The candidates are n8n (more control, steeper setup), Make.com (easier visual builder, recurring cost), or building most of it in code plus cron. I'll also teach this stack to a community of marketers new to AI, so beginner-friendliness matters too. Which do I commit to as the primary, and why?Sequencing the launch roadmap.
council this: I'm solo and capacity-constrained, and three big builds are competing for the same hours over the next two months: (1) launch weekly long-form YouTube, (2) pre-load my paid community curriculum so I don't open empty, and (3) start outreach to land my first case-study clients. All three matter for a September launch, but I can't do all three well at once. What do I sequence first, second, third, and what do I consciously delay?Sponsorship admin: VA, talent manager, or stay solo.
council this: I'm at 3 to 4 brand deals a month and the admin (contracts, invoices, negotiation back-and-forth) is eating my creative time. Option A: keep doing it myself and stay slow. Option B: hire a managed VA at a fixed monthly rate to handle admin only, while I keep all negotiation judgment. Option C: sign with a talent manager who takes 15 to 20% of deal revenue but owns the whole pipeline. Which structure do I commit to at my current deal volume?The Council is powerful, but it is not for everything. Skip it when: