Elizabeth forwarded five reels from the same account in a single batch and asked for an audit of everything Claude-specific on it. This is that report: what he posts, what actually makes it work, and five recommendations — each with its own verdict and, where approved, the written prompt that starts the work. The weekly inbox sweep is deliberately not here; it goes to Friday's Roundup.
Verified, 664K followers displayed, 405 posts. The bio is a funnel in three lines: "Founder @ Maker School · Helping 2000+ beginners land their first AI client · Get your first client in 90 days," pointing at a paid Skool community. Every follower count and student number here is his own displayed claim, unverified.
He matters twice. He is the clearest working example of the short-form engine our skit-show proposal wants to build — and Maker School is the productized teach-the-method business our own strategy defers to years two and three, running today. Inspiration and competition in one profile.
| Beat | What he posts | Reading |
|---|---|---|
| Claude Code & plugins | "Top 5 Claude Code Plugins" (twice) · "4 Plugins for Vibe Coding" · "5 Plugins to Stop AI Slop" · "Fix Your Claude Code Setup" · "Run Claude & Codex Together" · "Run Multiple AI Coding Agents" | His densest beat. Practical, repeatable, endlessly re-cuttable. |
| Cost & usage mechanics | "Fix Your Claude Usage Limits" · "Cut Your Claude Usage in Half" · "Run Unlimited Claude for Free" · "Unlimited AI Context Hack" · "Stop Paying for AI Tools" | Proven appetite for Claude operations. Also where his framing gets loosest. |
| Building & design | "Build $15,000 Websites with AI" · "Design Like Apple with Claude" · "Top 5 / Top 4 Claude Design Skills" · "Connect Claude to Any Tool" · "Build Mobile Apps with Claude" | Outcome-priced headlines on top of capability demos. |
| Odds and ends | "Free Claude Humanizer" · "Protect Claude from AI Virus" · "3 Best Claude AI Scrapers" · "Fable 5 Full Prompt Leaked" | The long tail, including the one item we should read and never imitate. |
Every thumbnail is the same object: one product logo across the top in white, two app icons flanking his face, a two-line benefit caption at the bottom with the second line knocked out in red. Same face, same crop, same type, hundreds of times. You can identify one of his posts at thumbnail size with the sound off and the title unread — which is the entire job of a feed image.
He is not winning on secrets; most of what he teaches is public documentation with a good headline on it. He is winning on a locked format applied relentlessly to a beat with real demand — a discipline rather than a trick, which is why it is honest to copy.
The opening is the half he cannot reach: he teaches Claude from the outside, and we run a fleet on it daily. Our machinery is the evidence his format cannot manufacture.
Cairo specifies the frame system — grid, type, subject and caption positions — tight enough that any agent hits it without a designer, and distinct enough that it is not the red-bar look every AI creator shares.
The finding above is that his format, not his content, is the engine. A locked system is the one part of that we can adopt honestly and immediately. It also has a deadline built in: the skit-show pilot should not generate a single frame before the system exists, or we spend the pilot retrofitting a look onto finished footage. Three worked examples and measured contrast come back as a proposal, not a deployment.
Awaiting both founders on this page.
Cairo, this week, ahead of any pilot frame being generated.
Ten to fifteen content angles on running Claude in production where the evidence is ours alone — real usage, real costs, real failures, real gate catches — each with a receipt and a named source artefact.
His cost-and-usage beat is his loosest framing and still one of his strongest performers, which means the demand is real even where the answers are folklore. Everyone in the category teaches Claude from outside it. We operate a fleet daily and can show the machinery — which is the same argument the Two-Chair positioning already makes about the business itself. This absorbs my existing T7 commitment rather than adding to it, and lands with the LinkedIn channel card that governs where the angles go.
Awaiting both founders on this page.
Aidan, by Wednesday 9 September, into Friday's Roundup.
Kai checks three recurring claims against Anthropic's own documentation — the ranked plugins, the "connect to any tool" surface, and the usage-limit folklore — and returns adopt, ignore, or already-have-it per claim.
Some of what he ranks may be a real capability gap in the Bench, and some of it is a confident video. We have been here before: the T1 pass found the lead-scraping reels oversold their method, and that finding is now load-bearing. The same discipline applies before any of this reaches a fleet skill file. Nothing recommended may put an agent outside its own gate.
Awaiting both founders on this page.
Kai, this week; same shape as the T1 pass.
Three genres run straight through his feed and all three are refusals for us: leak-framing, free-tier hacking, and unqualified outcome headlines.
Leak-framing — a post built on a leaked system prompt trades on someone else's disclosure failure, which is the exact currency an honesty-positioned agency cannot spend.
Free-tier hacking — "run unlimited Claude for free," "stop paying for AI tools" — is a vendor-relationship problem wearing a growth hook, and we are asking clients to trust how we handle exactly those relationships.
Unqualified outcome headlines — "$15,000 websites" with no engagement behind the number — is the claims ledger's whole reason for existing.
None of this makes him a bad source; it makes the boundary legible. The skip column is curriculum too, and it is cheaper to write down now than to argue in a comment thread later.
Nothing to build. A paid community teaching beginners to land AI clients, priced and populated, running at the market we plan to teach into in year two — worth knowing before that conversation arrives, not worth chasing now.
7. Research/Aidan_IG_Audit_nick_saraev_2026-09-06.md2. Business/Plan/HI_Strategy_TwoChair_2026-08-25.md §1 positioning, §6 product path.