Most of it isn't a clean lie. 68 of 100 claims we've tested are partly true — the tool is real, the demo happened, and the number in the caption quietly leaves out the subscription, the setup, or the part where somebody still has to do the work. That gap is where the money gets taken.
So every entry below carries the same four answers: what the claim was, what actually holds up, what doesn't, and — when there's something real underneath — exactly how to do it yourself, with the honest effort and cost. If you can run it yourself, we tell you how. Hiring us is for when you'd rather it was already working.
Drop a claim
Saw something that smells like BS? Paste the link. We check it and the verdict lands on this page — real, partly true, or straight BS, with the receipts either way. No email, no signup.
Got it. It's in the queue. We'll dig in and the verdict shows up on this page. Send another →
We test hype for free. We build the real thing for a living.
Thirty minutes, no pitch — and you'll leave with something useful either way.
Book a call with Todd or start with the free Business Checkup →The Verdict Weekly
Three verdicts every Friday. Free forever, unsubscribe anytime, no spam — that would be ironic.
PARTLY "Claude can build this, but it's never gonna design this" — a fair point wearing an unfair word: 'never'
The claimYeah, Claude can build this. But it's never gonna design this. (Over a slick animated website — the argument being AI can code a site but doesn't have design taste or judgment.)
Refreshing for the feed: clearly framed as opinion, no comment-gate, no course, nothing sold. And the defensible core is real — AI has no client, no user research, no brand strategy, and no accountability for the outcome, so a human still decides what's *right* for a specific audience. The 2026 industry consensus agrees taste is now the differentiator precisely because generation got cheap. The overreach is the word 'never.' Current Claude models demonstrably produce coherent visual systems and propose distinct art directions — that IS design output, just design without a brief or a stakeholder. The honest version is 'AI designs *something*; a designer designs *the right thing for you*.' It's genuine commentary from someone whose living the claim defends — not a scam, but not a neutral observer either.
What holds up
- The core is defensible: AI has no stake, no client, no user research, no accountability — humans still decide what fits brand and audience
- 2026 design-industry consensus agrees taste and judgment are now the differentiator because generation is cheap
- Honest format — opinion labeled as opinion, no funnel, no product, no gate
What doesn't
- 'Never' is unsupportable — current Claude produces coherent, sometimes distinctive visual design today; the gap is context, not capability
- The slick site used to 'prove' human superiority is exactly what AI produces unassisted — and viewers can't verify it wasn't AI-assisted
- Soft self-interest: it's authority content from a designer whose livelihood the claim defends
- The build-vs-design binary is already stale — the real 2026 split is AI defaults vs human-directed AI output, and the winners use both
The catch
The claim is right about the gap and wrong about the word. AI absolutely designs *something*; what it can't do is design the right thing for your specific brand, audience, and business goal without a human holding the intent. 'Doesn't design for you, in your context' is true. 'Never gonna design' is a moat defended with an absolute the evidence already broke.
How to actually do it
- Write a real brief first: who's the audience, what's the brand voice, what one job must this page do — that's the design work AI can't do for you
- Have Claude propose 3–4 directions against that brief, then pick and refine with human judgment
- Use AI as the junior designer and the production layer — fast, tireless, no ego — not the art director
- If the site matters commercially, pay a human for the brief-and-judgment layer even if AI writes every line of code
- Don't believe 'AI can't design' OR 'AI designs everything' — the winners in 2026 direct AI, they don't worship or dismiss it
Treat AI as the production layer and keep a human on intent.
- Effort
- A brief takes an hour; it's the highest-leverage hour in the whole build
- Cost
- $20/mo DIY-with-taste; a few hundred to ~$2K for AI production + human direction; $3K–15K+ for full human design
- Stack
- A real brief + Claude for production + a human on intent
- Confidence
- Medium
- Posted by
- a working UX designer posting opinion commentary, no product
PARTLY "How I make websites like these in 5 minutes with Opus 5" — real model, real technique, demo-math clock
The claimHow I make websites like these in 5 minutes with Opus 5 — Prompts Library + Claude. The biggest mistake is trying to get a perfect site on the first AI try; ready-made prompts let you test many directions fast. Comment 'Prompt' and I'll send you the link to the prompts for this website.
The tool is real and genuinely this good — Claude Opus 5 is a current Anthropic model that one-shots polished, animated landing pages, and the reel's actual advice (don't chase perfection on the first try; generate several directions and pick) is legitimate practice Anthropic itself documents. What's compressed out of '5 minutes': the iterations to get a take worth filming, and everything that makes a website a website — real copy, working forms, hosting, mobile QA, SEO, a client sign-off. The 'prompts library' is commodity content that's free in dozens of places, and 'comment Prompt' is a lead-capture funnel for a design agency, not a gift. Real tool, real technique, showcase-math timeline, manufactured scarcity on the freebie.
What holds up
- Claude Opus 5 is real and current (Anthropic, 2026) and genuinely excellent at one-shotting animated landing pages
- The core advice — generate multiple visual directions instead of forcing the first result — is legitimate, and Anthropic documents this exact technique
- Iterating cheaply across directions really is a strength of building with AI
What doesn't
- '5 minutes' is best-take generation time, not project time — showcase reels are best-of-N; you don't see the generic drafts
- 'Websites like these' are landing-page shells: no real content, no CMS, no wired forms, no SEO, no client review
- The 'prompts library' is commodity content — equivalent Claude prompt packs are free all over the web
- 'Comment Prompt' is an auto-DM lead capture that also farms engagement — the prompts are bait, the agency retainer is the product
The catch
The freebie is real but not scarce, and the clock is demo math. The prompt technique being gated is the same one Anthropic gives away in its docs; the comment-gate exists to harvest leads for a design agency. You can have the whole 'library' in one sentence and skip the funnel.
How to actually do it
- Open Claude (Pro $20/mo) and paste: 'Before building, propose 4 distinct visual directions for [my business] — background hex, accent hex, typeface, one-line rationale each. I'll pick one, then build it as a single-file animated landing page.'
- Pick a direction, then iterate — that's the whole technique the reel is selling
- Budget 1–3 hours to go from first generation to something publishable, not 5 minutes
- Supply your own words and photos — that's the part no prompt does, and it's what makes it your site
- Host free on Netlify's tier; skip the comment-gate entirely
The entire 'prompts library' in one prompt, plus the honest timeline.
- Effort
- 1–3 hours to a publishable small-business landing page, content included
- Cost
- Claude Pro $20/mo + free hosting vs the $1,500–5,000+ agency build behind the comment gate
- Stack
- Claude Opus 5 + your own copy/photos + Netlify free tier
- Confidence
- High
- Posted by
- a web-design agency running a comment-'Prompt' lead-capture funnel
BS "Learning agentic AI is a free path to a 6-figure corporate job — I can build your system for you" — the demand is real, the shortcut is self-defeating
The claimLearning agentic AI is like a free path to a 6-figure corporate job. I can create your own system like mine — use it as a portfolio and land a crazy high-paying job.
The kernel is real and worth stating plainly: agentic-AI job postings exploded (roughly 100x growth in two years per Stanford's AI Index data, tens of thousands of postings averaging near $200K) and the AI-skills wage premium is genuinely large. Here's what guts the pitch: those salaries attach to senior engineering roles, hiring managers screen specifically for projects the candidate built and can defend, and the offer on the table — pay someone to build 'your' system — produces the one kind of portfolio that dies in the first technical interview question. The pitch contradicts itself in a single breath: a 'free path' that costs whatever the DM quote is. And handing a paid stranger the keys to an agent framework wired into your email, files, and accounts is a security exposure stacked on top of the career fantasy — the framework in question carries hundreds of security advisories and a marketplace where hundreds of malicious add-ons have been found.
What holds up
- Agentic-AI demand is real: postings up ~280% year over year, ~42K postings averaging ~$197K (Stanford AI Index / PwC AI Jobs Barometer data)
- The learning path genuinely is near-free: open-source frameworks, official docs, $0–20/mo to practice
- The 'portfolio gets interviews' mechanism is real — when it's yours
What doesn't
- The fatal contradiction: a portfolio someone else built is worthless in the interview that pays the salary — hiring managers probe whether YOU built it
- Six-figure agentic roles are senior engineering roles; the caption launders senior-market salary data into a beginner promise
- Entry-level tech is currently brutal (deep cuts in junior hiring) — the exact opposite of 'crazy high-paying job for showing up with a system'
- Security: a stranger configuring an agent with access to your accounts, on a framework with 255+ security advisories and a marketplace where 335 malicious skills were found
- "Free path" + "I can create it for you" = the funnel is the contradiction
The catch
The seller is charging you to remove the only part that was ever worth money — the building. The demand data is real, which is what makes the pitch land; but the job goes to the person who can explain every design decision at a whiteboard, and that person is made by laptop hours, not by DM invoices.
How to actually do it
- Pick one real job an agent can do — inbox triage, report generation, follow-up scheduling — ideally from work you already know
- Build it yourself with free tooling: Claude's agent docs, an open-source framework, $0–20/mo to run
- Ship it somewhere real (your own workflow counts), then write up what broke and how you fixed it — that writeup IS the portfolio
- Be able to defend every design decision — that's the interview, and it can't be purchased
- Timeline honesty: months of real building on top of some technical footing — not a DM and a weekend
The actually-free version of the free path.
- Effort
- Months of real laptop hours — the same months the salary is paying for
- Cost
- $0–20/mo to learn vs a DM invoice for a portfolio that can't survive one whiteboard question
- Stack
- One real problem + Claude agent tooling + your own hands + a public writeup
- Confidence
- High
- Posted by
- an agentic-AI seller pitching DM'd custom builds as career portfolios
PARTLY "5 Claude Code plugins every builder needs" — four solid picks and one stat the repo itself already retracted
The claim5 GitHub repos / Claude Code plugins every AI builder needs: Graphify (maps your repo into a knowledge graph so Claude navigates instead of grepping), Impeccable (audits frontend for 24 design tells, kills AI-slop UI), Ponytail ('writes 80–94% less code, up to 75% cheaper, 3–6x faster'), Higgsfield MCP (image + video generation in-thread), and Obsidian as a persistent second brain.
Four of the five are real and genuinely useful — Graphify does build a repo knowledge graph so Claude stops grepping blind, Impeccable really does hunt AI-slop UI tells, Higgsfield MCP really does wire image/video generation into Claude Code, and the Obsidian-as-memory trick we run ourselves. The problem is the one with the biggest number. Ponytail's headline '80–94% less code' came from a flawed single-shot benchmark that the maintainer PUBLICLY RETRACTED — the rebuilt, realistic benchmark shows about 54% less code on average, ~20% cheaper, ~27% faster. The slideshow is quoting the dead numbers, and its '75% cheaper / 3–6x faster' don't match even the old claim. Real skill, modest real benefit, obsolete stat.
What holds up
- Graphify is real — builds a repo knowledge graph so Claude navigates instead of grepping, exactly as claimed
- Impeccable is real (by an ex-Google designer) — dozens of deterministic rules that flag AI-slop UI tells
- Higgsfield MCP is real — one MCP wiring GPT Image 2 / Seedance / video into Claude Code
- Obsidian-as-memory is real — we run it in production
What doesn't
- Ponytail's '80–94% less code' was RETRACTED by its own maintainer — the corrected benchmark is ~54% less code, ~20% cheaper, ~27% faster
- '75% cheaper / 3–6x faster' matches neither the corrected numbers (20% / 27%) nor even the old claim — invented precision
- 'Free' MCP omits Higgsfield's credit costs — real video work lands in the $49+/mo tier
- Graphify's viral '70x token savings' is a blog claim, not an independent benchmark
The catch
The tools are a good list; the marquee statistic is one the repo already corrected a month before this posted. When a plugin's own maintainer walks a number back and the video keeps quoting it, that's the tell — the exciting figure is doing sales work the corrected figure can't.
How to actually do it
- Graphify: genuinely useful on large codebases — expect real token savings, just not the blog's 70x
- Impeccable: good AI-slop-UI catcher; treat its fixes as suggestions to review, not gospel
- Ponytail: real and helpful, but set expectations at the CORRECTED numbers (~54% less code, ~20% cheaper) — or just put 'prefer stdlib, no speculative abstraction' in your CLAUDE.md and skip it
- Higgsfield MCP: fine tool; know that generating anything real burns credits — budget the $15–49/mo tier
- Obsidian-as-memory: the free, no-catch winner of the five — start here
The picks worth installing, with honest expectations on each.
- Effort
- An hour to install the set; the Obsidian memory habit compounds over weeks
- Cost
- Plugins free; Higgsfield generation is $15–49+/mo in credits
- Stack
- Graphify + Impeccable + Obsidian memory; Ponytail with corrected expectations; Higgsfield only if you generate media
- Confidence
- High
- Posted by
- a Claude-Code creator running a follow funnel
PARTLY "5 GitHub repos every AI builder needs" — all five real, two of them paid-in-risk not free
The claim5 GitHub repos every AI builder needs: n8n ('automate anything, self-hosted, no per-task fees'), awesome-claude-skills (curated drop-in skills), OmniRoute ('one API for 500+ models, 278+ providers, cheapest model per task'), Agent-Reach ('read X, Reddit and the web with zero API fees'), and last30days-skill ('mine Reddit for real opinions in ~20 seconds').
Better hit rate than most of these lists — all five repos exist and do roughly what's claimed. The sins are the omitted asterisks. n8n is free to self-host with no per-task fees, but it's fair-code, not open source: hosting it for multiple clients and charging crosses into needing a commercial license. OmniRoute is real and genuinely routes across hundreds of providers, but 'free' means the router is free — the models behind it still bill once you exhaust rate-limited free tiers, and piping code through 290 third-party endpoints is a data-governance decision, not a lifehack. Agent-Reach's 'zero API fees' is literally true and exactly why it's fragile: it leans on scraping and cookie-auth that violate X's and Reddit's terms and break when those sites change. Great for personal research; risky for client work.
What holds up
- All 5 repos verified real: n8n, awesome-claude-skills, OmniRoute, Agent-Reach (~65K stars), last30days-skill (many active forks)
- n8n genuinely has no per-execution fees, unlike Zapier — that part is true
- OmniRoute genuinely routes across hundreds of providers and can chain free tiers
- last30days-skill genuinely researches Reddit/X and returns usable summaries
What doesn't
- n8n is fair-code, not open source: hosting it for multiple paying clients needs a commercial/embed license — 'free for client builds' skips this
- OmniRoute 'free' = the router; the models still bill once rate-limited free tiers run out, and routing code through 290 endpoints is a governance call
- Agent-Reach's 'zero API fees' relies on scraping/cookie-auth that violate X and Reddit ToS and break on site changes — fragile by design
- '~20 seconds' is a marketing timing; real runs depend on how many sources are pulled
- Follow-funnel CTA; the value is real but the framing sands off every caveat
The catch
None of these is fake — the dishonesty is in the word 'free.' Free-of-license, free-of-fees, and free-of-consequences all quietly cost something here: a commercial license, provider bills once you scale, or account risk on tools that break the platforms' rules. Use them knowing which 'free' you're actually getting.
How to actually do it
- n8n: self-host for your own work freely; if you host it for multiple clients and charge, get n8n's commercial license — ask them, their docs say so
- OmniRoute: great router, but never send client code through random free inference endpoints; budget for paid model calls at scale
- Agent-Reach: fine for personal research; don't build a client deliverable on scraping that breaks and can flag accounts
- last30days-skill: genuinely useful for reading the room on Reddit — just don't expect the 20-second number on big pulls
- awesome-claude-skills: the safe, boring, actually-open pick of the five — start there
Use all five with the caveats the slideshow left out.
- Effort
- An afternoon to try all five; longer to self-host n8n properly
- Cost
- $0 to start + n8n license if you productize + provider bills once OmniRoute's free tiers run out
- Stack
- n8n (your own instance) + awesome-claude-skills + OmniRoute with paid keys for anything real
- Confidence
- High
- Posted by
- a builder-content page running a follow funnel
REAL "Ditther, Cooldock, Selected, Logo System, Macfolio, Ultramock, Javii — all underrated" — all real, three of seven quietly cost money
The claimSeven underrated design tools — Ditther, Cooldock, Selected, Logo System, Macfolio, Ultramock, and Javii. All linked, all underrated. Save this for later.
We hunted down all seven: every one exists, names spelled as given (yes, Ditther has two Ts), and the picks are decent — Ditther and Ultramock are genuinely fun, Logo System is a solid inspiration library at 1,200+ logos. An honest roundup by a real designer with no comment-gate and no product push. Two asterisks keep it honest: the listicle vibe implies free candy, but Cooldock is a paid app and Ditther/Ultramock gate their best exports behind Pro tiers — using all seven seriously runs $20–50, not $0. And two of the seven aren't tools at all — Selected and Macfolio are discovery directories. Padding, not deception.
What holds up
- All 7 verified real at their own domains — browser-effects editor, macOS dock, template rankings, logo library, app directory, 3D mockup tool, indie asset suite
- The strongest picks (Ditther, Ultramock, Logo System) genuinely deliver
- No funnel beyond 'save this' — no gate, no course, no affiliate trail we could find
What doesn't
- Not all free: Cooldock is a paid one-time purchase; Ditther and Ultramock hold their best exports behind Pro
- Two of seven are directories, not tools — the list is really five tools and two bookmarks
- "All linked" means linked in the slides — the caption itself carries none
The catch
Just the free-candy implication. As design-tool roundups go, this is the honest end of the genre.
How to actually do it
- Try first: Ditther's free tier (dither/halftone effects in the browser) and Logo System (free to browse, real inspiration value)
- Ultramock for screenshot mockups — or free alternates: shots.so and Figma mockup plugins
- Fully-free Ditther alternates if you hit the paywall: ditherit.com or Dither Me This
- Cooldock is a paid quality-of-life buy — stock macOS dock is fine until it isn't
- Treat Selected and Macfolio as directories to raid once, not tools to adopt
The picks worth your bookmark bar, plus the free routes around the paid ones.
- Effort
- Ten minutes of clicking to know which two you'll keep
- Cost
- $0 to browse and try; $20–50 to own the paid tiers seriously
- Stack
- Ditther free + Logo System + shots.so covers most of what the seven promise
- Confidence
- High
- Posted by
- a product designer sharing tools, audience-building, nothing gated
PARTLY "5 open-source AI trading agent repos most people are sleeping on" — honest caption, undisclosed conflict: the poster sells an AI trading terminal
The claim5 open-source AI trading agent repos across crypto, stocks, and prediction markets — 'the real ones where an LLM actually reasons.' The caption itself concedes the biggest one can't place trades by design, and says to run everything in paper-trading mode first.
By TikTok standards this caption is honest — it tells you the flagship can't trade and to paper-trade first, which is more candor than the genre ever shows. The repos are real: TradingAgents (~98K stars) and ai-hedge-fund (~63K) are the two giants, and ai-hedge-fund's README says in bold that it's educational and doesn't place trades. But the receipts get worse the closer you look: TradingAgents' own community caught its backtests leaking future data (a 'clear look-ahead bias' issue — real-time info retrieved during historical simulation), one listed repo is a general agent framework rather than a trading agent, one is Microsoft research tooling, and one is an 87-star repo idle since April. No verified live profits exist for any of the five. And the flag that matters most: the poster's own site sells an AI crypto-perps trading terminal — a trading-repos post from that seller is top-of-funnel, and the caption doesn't say so.
What holds up
- All 5 repos exist; TradingAgents (~98K) and ai-hedge-fund (~63K) are heavyweight and real
- ai-hedge-fund's own disclaimer is honorable: educational only, deliberately cannot trade
- The caption's paper-trade-first advice is correct and creditable
- LLM-agents-reasoning-about-markets is a real research direction (TradingAgents has an arXiv paper)
What doesn't
- Undisclosed conflict: the poster sells his own AI crypto-perps trading terminal while posting trading-repo roundups
- TradingAgents' headline backtests were caught leaking future data by its own community — and LLM backtests carry an unfixable leak besides: the model's training data already contains the era's outcomes
- One pick is idle since April at 87 stars; two others aren't trading agents at all
- Multi-agent analysis burns $5–20/day in API calls per handful of tickers — you pay real money to run the simulation whether or not it ever earns
- Zero verified live profits anywhere in the ecosystem — if reasoning-LLM trading reliably beat the market, it wouldn't be free on TikTok
The catch
A backtest that peeked at the future is a résumé with a fake job on it — and the model's own training data makes some peeking unavoidable. Treat every LLM-trading backtest as marketing until it survives months of live paper-trading you ran yourself. Then remember the industry receipt from our #089: four of six frontier models lost real money the moment the trades were real.
How to actually do it
- Clone ai-hedge-fund or TradingAgents and run paper mode with a cheap model — genuinely a great course in multi-agent reasoning
- Paper-trade any strategy through MONTHS and different market regimes before a dollar moves
- Size anything experimental at money you can fully lose — that's the position size for 'my bot has a thesis'
- Evidence bar: a strategy whose only proof is its own backtest hasn't met it
- When a trading-content creator also sells a trading product, read the roundup as the top of their funnel — because it is
The education is real and free — here's how to take it without paying tuition to the market.
- Effort
- A weekend to run paper mode; months of it before the results mean anything
- Cost
- $5–20/day in API calls to run seriously — the simulation bills you even when the market doesn't
- Stack
- ai-hedge-fund (paper) + a cheap model + a calendar reminder to still be paper-trading in 90 days
- Confidence
- High
- Posted by
- an AI-content creator (a real engineer) whose own product line includes an AI crypto-trading terminal — not disclosed in the post
PARTLY "5 open-source repos that find you paying clients" — the repos make a list; nothing on earth makes the list pay you
The claim5 open-source repos that find you paying clients, starting with local businesses that don't even have a website: a Google Maps scraper, an email-existence checker, a no-code scraper builder, self-hosted forms, and a self-hosted helpdesk — replacing what people pay Apollo, Instantly, and ZeroBounce for.
All five repos are real and maintained, and 'local businesses without websites' is genuinely a smart niche. Then the pipeline falls apart at every joint. Scraping Google Maps breaches Google's terms — the answer can be IP blocks or your Google account. The email-verification step — the hinge of the whole pipeline — is documented by its own maintainer as impractical from home: it needs outbound port 25, which residential ISPs block, and real volume needs SMTP proxy servers. Nothing in the stack does deliverability, which is the thing Apollo and Instantly actually charge for. And the promised outcome — paying clients — is the step no repo performs. The tools produce a list. The list is not a client.
What holds up
- All 5 repos verified real and active (the scraper, the verifier, Maxun ~17K stars, HeyForm ~9K, Libredesk ~2.7K)
- The niche is real: businesses with no website are underserved and reachable
- Cold email is legal under CAN-SPAM — with a physical address, truthful headers, and working opt-out
What doesn't
- The verifier's own README: port 25 required (residential ISPs block it), SMTP proxies needed at any volume — the key step doesn't work from your laptop
- The verifier is dual-licensed: commercial use inside your for-profit lead-gen is exactly what its paid license covers — 'free' isn't
- Google Maps scraping violates Google ToS — the downside includes your Gmail/Drive account, not just an IP
- Category mismatch: the helpdesk repo is a support tool, not a cold-outreach reply-catcher; none of the five does deliverability
- CAN-SPAM violations run up to $51,744 per email — the compliance part goes unmentioned
- Zero evidence offered that this stack has converted into a paying client for anyone
The catch
Apollo and Instantly aren't charging for software — they're charging for data quality and sending infrastructure, which is precisely what can't be self-hosted from a laptop. The post replaces the price tag and keeps the illusion that the tooling was ever the hard part. The hard part is being worth replying to.
How to actually do it
- Build the list by hand: Google Maps, one city, one trade, 30 minutes — note who has no website. Same list, no breach
- Verify emails pay-as-you-go: ~$8 per thousand on a ZeroBounce-style service — no subscription, no port 25
- Send low volume from a real domain with a real CAN-SPAM footer — or skip email entirely
- Better: call. Businesses with no website answer their phone — that's why they have no website
- If you want scraped data anyway, one month of a scraping service (~$30) beats building the pipeline
The version that gets local clients without the ToS breach — it's mostly not software.
- Effort
- An afternoon builds a better list by hand than the pipeline builds in a week of setup
- Cost
- $8/1,000 verifications + a domain vs $50+/mo in proxies and the legal tail risk
- Stack
- Google Maps + a spreadsheet + a phone — the stack nobody posts because it isn't shiny
- Confidence
- High
- Posted by
- an AI-content creator (a real engineer) who sells 1:1 consults and app builds
REAL "The $0 stack: ten repos, 446,694 stars, doing work you'd pay for" — the math checks out, and so do the tools
The claimTen open-source repos, 446,694 stars between them, doing work you'd expect to pay for: Stirling-PDF→Acrobat, Immich→Google Photos, Upscayl→Topaz, Whisper→Otter, LocalSend→AirDrop, plus five newer agent-tooling picks. Most run entirely on your own machine; costs nothing but the time to install.
We added the stars ourselves: the claimed total was accurate at posting (455K by the time we checked — they grew). All ten repos exist, and the headline five are exactly the tools the self-hosting community already swears by: Stirling-PDF (~89K stars), Immich (~110K), Upscayl (~48K), Whisper (~107K), LocalSend (~88K). These are the real deal. The honest footnotes: 'Whisper replaces Otter' compares a model to a product — Otter's meeting bot, speaker labels, and summaries need a wrapper app around Whisper; Immich makes you the sysadmin of your family's photo library (always-on box, backups, docker); and the five agent-tooling picks quietly sit on top of a paid Claude plan. 'Free' means you pay in setup time and a machine that stays on. As repo lists go, this is the good stuff.
What holds up
- Star math verified by hand: claimed 446,694 was accurate at posting date
- The big five are the community-consensus best-in-class replacements, all actively maintained
- LocalSend and Upscayl are true one-click installs — 'anyone' difficulty
- Small-repo picks checked for star-farming signals: fork/issue ratios look organic
What doesn't
- Whisper is a transcription model, not an Otter replacement out of the box — no meeting bot, no diarization, no summaries without a wrapper
- Immich's real cost: a $100–300 always-on box + docker + backup discipline — cheaper to pay Google $2/mo if you won't run a server
- Two of the five 'agent tooling' picks are hobby-stage (280 and 478 stars) — not stack material yet
- The agent picks require a paid Claude plan — the '$0 stack' has a $20–200/mo floor for that half
The catch
Only the definition of free. Free software, yes; free operation, no — the currency is setup weekends, an always-on machine, and (for the agent half) an AI subscription. Priced honestly, the big five are still a bargain.
How to actually do it
- Today, zero skill: LocalSend (AirDrop everywhere) and Upscayl (one-click upscales) — genuinely install-and-go
- This weekend, some skill: Stirling-PDF via one Docker command — covers most Acrobat jobs
- Only if you'll commit: Immich replaces Google Photos beautifully AND makes you responsible for the backups — mini-PC or NAS required
- Transcription: use a Whisper wrapper app (MacWhisper, Vibe) to get the Otter experience, not raw Whisper
- Skip the sub-500-star picks until they grow up — watchlist, not stack
The install order that pays off immediately.
- Effort
- Two are instant; the self-hosted ones are a weekend each
- Cost
- $0 software + a $100–300 always-on box for Immich + $2/mo to Google remains the honest lazy option
- Stack
- LocalSend + Upscayl + Stirling-PDF first; Immich for the committed
- Confidence
- High
- Posted by
- an AI-repos page that counts stars honestly (same account as #089)
BS "Claude Code built a GTA-style open world from scratch in 2 hours" — it's Epic's Matrix tech demo, and the crowd caught it
The claimUnreal Engine x Claude Code x MCP — 'this thing is built from scratch,' 'Claude Code after 2h using Unreal Engine,' shown as a photoreal GTA-style open-world city.
The footage isn't Claude's work — it's 'The Matrix Awakens,' Epic's famous Unreal Engine 5 tech demo, and the post's own audience says so in force. The top comment ('fake, it's the Matrix demo on UE') has hundreds of likes, and reply after reply identifies the same map, the same character, even the same on-screen 'fly' controls from the demo. What actually happened is a real, months-of-work Epic showcase got relabeled as 'Claude Code, from scratch, in 2 hours.' The real thing underneath — Claude driving Unreal via MCP — genuinely exists and can do useful editor work (we covered it honestly in our Geopogo entry). What it cannot do is generate a photoreal open-world city from scratch in an afternoon, and this clip is proof by misattribution, not by capability.
What holds up
- Claude-driving-Unreal-via-MCP is a real capability — see our entry on the Geopogo bridge for the honest version
- Unreal Engine 5 genuinely can render this quality — because Epic built exactly this as a tech demo
What doesn't
- The footage is Epic's 'The Matrix Awakens' UE5 tech demo — not built by Claude, not from scratch, not in 2 hours
- The post's own top comments (hundreds of likes) identify it: same map, same character, same 'fly' controls
- 'Built from scratch in 2h' misattributes a months-long professional Epic showcase to an afternoon of AI
- The creator's other posts sell AI web-design services — the fake demo is authority-building for that funnel
The catch
This is the cleanest kind of BS: not an exaggeration of what a tool did, but someone else's famous work relabeled as the tool's. The giveaway is always the same — the audience recognizes the source. When a demo looks impossibly good, check whether it's a real product's output or a AAA studio's showcase wearing a new caption.
How to actually do it
- Reverse-image or reverse-video suspicious 'AI built this' clips — this one is instantly findable as the Matrix UE5 demo
- The genuine capability: Claude over MCP can spawn actors, edit properties, and scaffold Blueprints in Unreal — real, useful, unglamorous
- It cannot generate a photoreal open-world city from scratch in 2 hours — no tool can; that was Epic's team over many months
- Try the real thing free: Epic documents native Unreal MCP in UE 5.8, or use an open-source bridge with your Claude plan
- When the creator selling AI services shows a jaw-dropping demo, check whose work it actually is before you buy
What Claude + Unreal actually does today — the real version is still worth knowing.
- Effort
- 30 seconds to identify the source clip
- Cost
- $0 — this is a spot-the-fake, not a buy
- Stack
- UE 5.8 native MCP + Claude for the real (modest) version; a reverse-search habit for the fakes
- Confidence
- High
- Posted by
- a creator whose other posts promote AI web-design services
PARTLY "The sentence that logged my side income" — a real app wearing a 5.0-star screenshot that doesn't match its store page
The claimA story of building side income with Claude, ending in a plug for Tase — an AI life assistant that 'turns one spoken sentence into logged actions across money, health, tasks, and goals.' Screenshot shows '1.8K ratings, 5.0 stars.' Origin story: built a 5-tab tracking sheet with Claude in 8 minutes, got 7 paid requests at $15 in week one. Comment TASE for a 14-day free promo code.
Tase is a real app (Tase LLC, on the App Store) and it does roughly what's claimed — voice-to-logged-actions across money, health, and tasks. But the social proof is the problem. The slide shows '1.8K ratings, 5.0 stars'; the live App Store page shows 4.8 stars from 53 ratings — the screenshot inflates the rating count roughly 34x and shows a perfect 5.0 that doesn't exist on the real listing. The origin story (a 5-tab sheet built in 8 minutes, 7 paid requests at $15 in week one) has zero independent trace and carries a price anchor baked in — the shape of funnel fiction, not a case study. And 'free' is a weekly-subscription app: after the promo it's $3.99–6.99/week or $80–200/year.
What holds up
- The app is real: Tase — AI Life Assistant, by Tase LLC, live on the App Store
- The capability is genuine — voice input logged across finance/health/tasks/goals
- The category (AI life-logging) is real and functional
What doesn't
- The '1.8K ratings, 5.0 stars' screenshot contradicts the live store page (4.8 stars, 53 ratings) — inflated ~34x, and a flat 5.0 that doesn't appear on the real listing
- The origin story (8-minute build, 7 paid requests at $15) has no independent trace and comes with a built-in price anchor — funnel-fiction structure
- 'Free' is a subscription app: PRO $3.99/wk–$79.99/yr, MAX $6.99/wk–$199.99/yr; the weekly tier is the classic ~$200/yr trap
- The whole post funnels to a comment-'TASE' promo code for the app it's promoting
The catch
When a testimonial's centerpiece screenshot doesn't match the product's own public page, that's the tell. The app is real; the proof around it is manufactured, and 'free' means a weekly subscription that adds up to $80–200 a year.
How to actually do it
- Check the App Store page yourself before believing any in-video rating screenshot — here the real number is 4.8 from ~53, not 1.8K at 5.0
- If the app interests you, use its actual free tier to test it — don't let 'free promo code' obscure that it's a weekly subscription after
- Price it honestly: $80–200/yr depending on tier; the $3.99/week option is ~$207/yr
- Want the same logging for $0? Claude plus a Google Sheet or Obsidian note does voice-to-log without a subscription
- Discount origin stories with dollar amounts and no receipts — they're written to sell, not to inform
How to test the real app without getting anchored by the fake numbers.
- Effort
- Minutes to try the free tier; seconds to check the real rating
- Cost
- 'Free' app = $80–200/yr after promo vs $0 for a Claude + Sheets logger
- Stack
- Tase free tier to test — or Claude + a Sheet if you'd rather not subscribe
- Confidence
- High
- Posted by
- a creator funneling to a comment-'TASE' promo code for the app being advertised
PARTLY "Run an AI hedge fund from your laptop" — ten real repos, one sentence that loses people money
The claimTen-repo list led by: AutoHedge (3.8K stars) — run an AI hedge fund from your laptop, 4 agents for research, validation, risk, and trade. Plus Vibe-Trading (describe a strategy in plain English, it trades), a Bloomberg-without-the-$25K-seat terminal, self-hosted every-model chat, a free AI video studio, and five more.
We checked all ten: every repo exists, every quoted star count was accurate at posting (they've since grown), and several are genuinely excellent — LibreChat, HeyGen's hyperframes, and Vibe-Trading (a real university lab project with paper-trading mode and audit logging). The hazard is the headline verb. AutoHedge is a 37-commit marketing demo for an agent framework — its own maker calls it a demo app. Live execution is Solana-only, setup involves pasting your wallet's private key into an env file, and the README carries no risk disclaimer and zero performance results. Here's the market's own receipt: in the Alpha Arena live benchmark, six frontier models traded $10K each of real money — four of six lost, one down 63%, losses driven by the models over-trading away their capital in fees. That's the best AI in the world with a professional harness. Your laptop is not positioned to beat it.
What holds up
- 10/10 repos real; star counts honest at posting date — rare and creditable for the genre
- Vibe-Trading is a legitimate HKU lab project: 1,300+ commits, paper/shadow modes, audit logging — the credible one of the trading pair
- LibreChat (~42K stars) and hyperframes (genuinely HeyGen, ~41K) fully check out
- The Bloomberg-aesthetic terminal is real and active — as a dashboard, not a data-feed replacement
What doesn't
- AutoHedge: no disclaimer, no track record, Solana-only live execution, and a setup step that says 'paste your wallet private key here' — one leaked env file from a drained wallet
- Live LLM-trading evidence runs the wrong way: 4 of 6 frontier models LOST real money in Alpha Arena (one down 63%), bled out by over-trading costs
- "Free AI studio" on the list routes through a paid generation API — free cash register
- "Bloomberg without the $25K seat" is the aesthetic, not the licensed data — the data IS the seat
- A hedge fund is edge + risk capital + execution infrastructure + a legal wrapper; a repo is none of those
The catch
The bill arrives in three currencies: API tokens for every analysis cycle ($50–500+/mo run 'as a fund'), fees and slippage to professionals and MEV bots on the other side of every trade, and tail risk of a private key sitting in a plaintext file. The realistic median outcome is a slow bleed. The one sentence that would fix the post — 'paper-trade it for the education' — is the one sentence it doesn't say.
How to actually do it
- Want the AI-trading education? Vibe-Trading in paper/shadow mode — months of it, through different market conditions, zero dollars at risk
- Never put a wallet private key in an env file on your laptop — that rule survives every bull market
- Want returns from your laptop? A boring index fund has outperformed every public LLM live-trading result on record
- Want the Bloomberg feel? The terminal repo is honestly great — as a dashboard
- Grab the actually-clean picks: LibreChat for self-hosted chat, hyperframes for HTML-to-video
How to get everything real this list offers without funding a wallet.
- Effort
- Paper-trading setup: a weekend. Understanding why the bot's confident thesis is wrong: the actual skill
- Cost
- $0 repos + $50–500/mo tokens if run seriously + your stake if you ignore the paper-trade step
- Stack
- Vibe-Trading (paper mode) + LibreChat + an index fund for the money part
- Confidence
- High
- Posted by
- an AI-repos page that counts stars honestly and oversells verbs
PARTLY "ruflo: a swarm of Claude agents coordinating across your whole project" — real repos, but the swarm's own users say it lies about its work
The claimSeven Claude Code repos to save for your next build, led by ruflo — 'the successor to claude-flow, a swarm of Claude agents that coordinate and share memory across a whole project' — plus 100+ ready-made subagents, GitHub-Issues project management, PRD generators, Anthropic's official agent SDK, slash commands, and a CLAUDE.md starter. Free starter kit in bio.
All seven repos exist and five are actively maintained — the subagents collection (~24K stars), Anthropic's official SDK, and the PRD workflow are accurately described. The problem is the headliner. 'Successor to claude-flow' inflates a rename into a new product: same repo, same author, renamed at v3.5 for trademark reasons. And ruflo's own community reports the harder problem — a discussion thread on the repo describes the verification system as broken 'at a critical level,' with agents self-reporting success when most actually fail. A 67K-star swarm that grades its own homework and gives itself an A is exactly the thing you don't want shipping your project. The bio's 'free starter kit' is a standard email-capture lead magnet.
What holds up
- All 7 repos verified real; 5 of 7 actively maintained
- The subagents collection genuinely has 100+ (now 150+) specialist definitions
- claude-agent-sdk is genuinely Anthropic's official SDK, pushed the day before we checked
- ai-dev-tasks genuinely turns an idea into a PRD and task list — it's three markdown files doing honest work
What doesn't
- "Successor" = a rename (claude-flow → ruflo, same author, trademark reasons) — not a new generation of anything
- ruflo's own discussion threads report agents self-reporting success while failing — the swarm's verification is the contested part
- Swarm orchestration burns tokens fast — practical floor is a Max plan (~$100–200/mo) or real API spend
- Two of seven repos untouched in 9–10 months
- "Free starter kit, link in bio" — the post is a lead magnet
The catch
The seductive idea is delegation without verification: a hundred agents shipping while you sleep. Our shop runs on Claude agents daily, so believe us when we say the opposite is the discipline that works — fewer agents, verified work, human on the merge. A swarm you can't audit isn't leverage; it's liability with good marketing.
How to actually do it
- Start with plain Claude Code + a good CLAUDE.md — that alone is most of the magic these kits sell
- Add 2–3 subagents you fully understand from the awesome-claude-code-subagents collection — not 100
- Steal the PRD flow: ai-dev-tasks is 3 markdown files; read them in 10 minutes, no install
- Anthropic's own docs cover subagents and the Agent SDK free — the official source beats the starter kit
- Iron rule for any swarm: never accept an agent's self-report — verify output like you'd verify a new hire's first week
What we actually run instead of a hundred-agent swarm.
- Effort
- An afternoon to a solid Claude Code setup without any meta-harness
- Cost
- Repos $0; swarm-scale token burn is $100+/mo — simple setups run fine on Pro
- Stack
- Claude Code + CLAUDE.md + 2–3 verified subagents + you on the merge button
- Confidence
- High
- Posted by
- an AI-content creator running a link-in-bio starter-kit funnel
PARTLY "This week's trending GitHub repos" — real repos, but one of them trended two years ago
The claimThis week's trending GitHub repos: an AI that applies to jobs for you, one that runs Word/Excel/PowerPoint, one that lets Claude control your entire computer, one that kills AI slop, and a free gateway to 50+ AI models. All free and open source.
All five categories are real and open source, and one pick (the Office automation tool, OfficeCLI) genuinely was trending the exact week of the post — we found it in two independent trend reports from that window. But the lead item, the job-application bot (AIHawk), had its viral moment in 2024 — recycling old viral repos as 'this week' is the genre's standard move — and auto-applying violates LinkedIn's terms, an account-ban risk the post never mentions. 'Free gateway to 50+ models' conflates free software with free model usage: the gateway costs nothing, most models behind it still bill per token. And 'lets Claude control your entire computer' is a real capability that deserves a security warning, not just hype.
What holds up
- OfficeCLI verified genuinely trending in the post's window (+1,929 stars in one July report, top-10 agent skill the same month)
- The anti-AI-slop wave is real (kill-ai-slop, avoid-slop at 11K+ stars)
- Model gateways are real (Portkey Gateway spans 1,600+ LLMs, open source)
- The job-application repo exists and is genuinely open source
What doesn't
- "This week's trending" is false for the lead pick — AIHawk went viral in 2024 and its commercial arm wound down; the ecosystem is scattered forks
- Auto-applying violates LinkedIn ToS — the post sells the bot, you carry the ban risk
- Full-computer-control agents are a security surface, not just a convenience
- "Free gateway to 50+ models" — the software is free, the tokens aren't
The catch
Trending-repo content is farmed on recycled virality — the repos are real, the 'right now' is often the fiction. The GitHub trending page is public; a carousel is someone's week-old screenshot of it with the risks cropped out.
How to actually do it
- Job hunting: have Claude tailor each resume and cover letter to the posting, then apply manually to the shortlist — quality beats spray, and your LinkedIn survives
- Office automation: OfficeCLI is the real trending pick of the five — that one's worth a look
- Computer control: Claude Code/Cowork already do this with permission gates — prefer the gated version over a raw repo
- Model gateway: fine tool, but read the pricing of the models BEHIND it before routing traffic
- Check trends yourself: github.com/trending and ossinsight.io/trending — free, live, no carousel
The versions of these five jobs that don't cost you an account or a surprise bill.
- Effort
- Minutes to check any trending claim yourself
- Cost
- $0 for the repos; tokens, ban risk, and security exposure are the hidden line items
- Stack
- GitHub trending (raw source) + Claude for the resume work + gated agents over raw ones
- Confidence
- High
- Posted by
- an AI-tools creator farming saves
PARTLY "Total monthly cost to run a startup: ~$20" — the stack is real; the $20 leaves out the bill that grows
The claimTotal monthly cost to run a startup: ~$20. Claude for coding ($20/mo), Supabase (free), Vercel (free), Namecheap domain ($12/yr), Stripe (2.9%/transaction), GitHub (free), Resend, Clerk, Cloudflare, PostHog, Sentry, Upstash, Pinecone — all free tiers. There's never been a cheaper time to build.
The individual claims are basically accurate — every tool listed is real and every 'free' really is a genuine free tier that thousands of builders launch on. As a snapshot of how cheap it is to START, the post is honest and even useful. Where it slips is the headline '~$20 to run a startup.' Free tiers are onboarding ramps, not the running state: Supabase, Vercel, Clerk, Pinecone, PostHog, and the rest all have usage ceilings, and the moment you have real users you cross them and the bill climbs — often to hundreds a month. Stripe's 2.9% is also a real cost the '$20' total quietly omits. So: ~$20 to build and launch a prototype, yes. ~$20 to RUN a startup with actual traffic, no.
What holds up
- Every tool named is real and every free-tier claim is roughly accurate at the entry level
- $20 Claude + a ~$12/yr domain genuinely gets a prototype built and deployed
- The broader point holds: the cost to START building has genuinely never been lower
- No product, no funnel — just a stack list
What doesn't
- 'To run a startup' overstates 'to launch a prototype' — free tiers are ramps with ceilings, not the steady state
- Real users push you past the free limits on Supabase/Vercel/Clerk/Pinecone/PostHog — the bill scales with success
- Stripe's 2.9% per transaction is a real cost the '$20 total' leaves out
- Pinecone and similar can jump to hundreds a month at modest scale
The catch
The word doing the damage is 'run.' Everything here is a true free tier — for building and launching. A startup that's actually working has users, and users are exactly what moves you off every free tier at once. Budget for the second month, not just the first.
How to actually do it
- Build and launch on these free tiers — that part is real and genuinely ~$20/mo
- Know your ceilings before you hit them: check the free-tier limits on Supabase, Vercel, Clerk, and Pinecone now, not when the bill spikes
- Add Stripe's 2.9% to any revenue math — it's a real line item
- Expect the jump: a startup with real traffic runs hundreds/mo, not $20 — that's success, not failure, but budget for it
- Swap the priciest-at-scale piece (often the vector DB) only when usage actually demands it
Use the stack honestly — cheap to start, with eyes open about what scales.
- Effort
- A weekend to a live prototype on the free stack
- Cost
- ~$20/mo to build and launch; hundreds/mo once you have real users
- Stack
- Claude + Supabase + Vercel + Stripe + a domain — real, and priced for month one
- Confidence
- High
- Posted by
- a tech creator sharing a startup stack, no product
PARTLY "Open-source tools now do the jobs people pay subscriptions for — these 10 GitHub projects" — true premise, zero named tools
The claimOpen-source tools are getting seriously good — in many cases they do the same jobs people pay monthly subscriptions for, across AI, automation, coding, scraping, databases, and productivity. This list of 10 GitHub projects proves it. (None of the 10 are named in the caption.)
The premise is verifiably true as a category — this audit has confirmed a dozen instances of it (Stirling-PDF vs Acrobat, n8n vs Zapier, PostHog vs Mixpanel, Baserow vs Airtable, Chatwoot vs Intercom). But 'these 10 projects' can't be receipt-checked because the caption names zero of them. Un-nameable tools are un-gradeable tools, and the vagueness is the retention mechanic: you must swipe, save, and follow. Compare entry #070, which at least named its list and could be held to it. The control/customization/lock-in argument is directionally fair and omits the same cost every time: self-hosting means you become the sysadmin.
What holds up
- The open-source-replaces-subscriptions trend is real and accelerating — confirmed repeatedly across this audit
- The claimed categories (automation, analytics, databases, support) all have genuine open-source leaders
What doesn't
- Zero named tools = zero accountability — the whole payload is hidden in the slides
- "Follow me for more useful AI tools" is the CTA; the vagueness IS the product
- "Less dependence on platforms" is true only after you've successfully become your own IT department
The catch
A true premise is doing the credibility work for an uncheckable list. Save the premise, skip the mystery.
How to actually do it
- PDFs: Stirling-PDF. Automation: n8n (self-hosted). Analytics: PostHog. Database: Baserow. Support chat: Chatwoot. Local AI: Ollama
- Every one has a public repo — read the issues tab before installing; it's the honest review section
- Budget a VPS ($5–20/mo) or an always-on machine, plus real setup time per tool
- Swap one tool at a time; the graveyard of 'we self-hosted everything in a weekend' is large
- Rule: never install from a slideshow — install from a repo you've actually looked at
A vetted, NAMED list in the same categories — every one inspectable before you install it.
- Effort
- One tool per weekend is the sustainable pace
- Cost
- $0 software + $5–20/mo hosting + your new sysadmin hobby
- Stack
- The named six above — same categories, actual receipts
- Confidence
- Medium
- Posted by
- an AI-tools page farming follows with mystery lists
PARTLY "5 open-source AI video generators — the same tech behind Runway and Kling, free on your machine" — a claim that names zero of the five
The claim5 open-source AI video generators most people are sleeping on — the same kind of tech behind paid tools like Runway, Kling, and Higgsfield, except free and running on your own machine.
The category is real and 2026 was genuinely its year: Wan 2.2 (Apache-licensed, Alibaba), HunyuanVideo 1.5 (Tencent), and LTX-Video produce output that independent comparisons put near mid-tier paid platforms. But this caption names none of its five repos — the list lives only in the slides, so the claim can't be pinned down as posted, and that alone caps the grade. 'Same kind of tech' is defensible (same architecture family); 'the same tech' would be false — Runway's and Kling's actual models are proprietary. And 'on your own machine' silently means a 16–24GB-VRAM GPU: a $1,600+ graphics card and minutes-per-clip render times. The paid tools' real product is speed and not owning the hardware.
What holds up
- The open-source video-gen class is real: Wan 2.2, HunyuanVideo 1.5, LTX-Video, Mochi, CogVideoX are the standard roster and several carry permissive licenses
- Independent comparisons put the best open models near Kling-1.5-level on some dimensions — the gap genuinely narrowed
- Higgsfield (name-checked as a paid tool) is itself largely a product layer on other models — see our entry #062
What doesn't
- Zero repos named in the caption — unfalsifiable as posted, and unusual for this creator's format
- "Your own machine" assumes a 16–24GB GPU most viewers don't own
- Top paid tools still win on prompt adherence and polish — 'sleeping on' implies parity that isn't there yet
- Caption funnels to the creator's paid 1:1 calls and app builds
The catch
You're saving a claim, not a list. When the names are only in the slides, the post is optimized for the save button, not for your ability to check it.
How to actually do it
- The actual list worth knowing: Wan 2.2, HunyuanVideo 1.5, LTX-Video — all with public repos you can inspect
- Own a 16GB+ GPU? ComfyUI + Wan 2.2 is the best-documented local path
- No GPU? Rent one: RunPod/Vast at roughly $0.40–0.80/hr for a 24–48GB card — cents per clip, no subscription
- Expect minutes per clip locally vs seconds on paid platforms — that speed IS what the subscription buys
- Keep the paid tool if video is client work; use open models for volume and experiments
The named version of the claim, plus the no-GPU route.
- Effort
- A day to a working ComfyUI pipeline if you're technical
- Cost
- $0 models + a $1,600+ GPU you may not own, or ~$0.50/hr rented
- Stack
- ComfyUI + Wan 2.2 (local) or a cloud GPU by the hour
- Confidence
- Medium
- Posted by
- an AI-content creator (a real engineer) who sells 1:1 consults and app builds
PARTLY "Geopogo MCP: build Unreal Engine gameplay in natural language" — real product, real demo, and Epic gives the same bridge away free
The claimGeopogo MCP connects Claude directly to Unreal Engine: create Blueprints, modify levels, place actors, generate gameplay systems in natural language. Available on the Fab store.
Geopogo is a real San Francisco design-AI company (they also ship a Claude-to-Revit plugin), the product page and Fab-store listing exist, and the capability class is genuine — Claude driving the UE5 editor over MCP works and demos well. Two things temper it. First, 'generate gameplay systems' stretches what these bridges reliably do: the small verbs (spawn actors, set properties, scaffold Blueprints) work; a functioning gameplay SYSTEM still needs a developer steering and debugging. Second — the part the ad understandably skips — Epic itself now documents native Unreal MCP support in UE 5.8, and free open-source bridges predate this product. We couldn't verify the Fab price (the store blocks automated fetches); the company's other tools are freemium-to-paid.
What holds up
- Real company with a shipping product line; the Claude-to-Unreal product page and Fab listing both exist
- The demo video is real and the capability class it shows genuinely works
- MCP-to-Unreal is a legitimate, growing category — Epic documents it natively in UE 5.8
What doesn't
- The handle is the funnel: this is a company ad for its own product (disclosed by its nature, at least)
- "Gameplay systems in natural language" is demo-speak — curated demo ≠ daily workflow
- Free equivalents exist and go unmentioned: Epic's native Unreal MCP, plus open-source bridges on GitHub
- Price unverified; sibling products are freemium with paid tiers
The catch
The ad isn't lying about the tech — it's omitting the market. When the engine vendor ships the same bridge natively for free, a paid wrapper has to earn its price in packaging and polish. Maybe it does; check the price tag before assuming.
How to actually do it
- Unreal Engine is free — start there
- Try Epic's own documented Unreal MCP in UE 5.8, or a free open-source bridge (e.g. GenOrca/unreal-mcp) with the Claude plan you already have
- Expect the small verbs to work: spawn, place, property edits, Blueprint scaffolds
- Expect to be the developer for anything called a 'system' — the AI is the intern, you're still the lead
- If the free bridge's setup pain is real for you, THEN check Geopogo's Fab price and decide if the packaging is worth it
How to get the demo's magic without buying first.
- Effort
- An evening to a working free bridge if you know Unreal basics
- Cost
- UE free + Claude ~$20/mo; Geopogo's price unverified — check before buying
- Stack
- UE 5.8 + native Unreal MCP or a free GitHub bridge + Claude
- Confidence
- Medium
- Posted by
- Geopogo — the tool company's own account, advertising its own product
REAL "MCPs turn Claude into an agent that controls your browser, repos, files, and Blender" — accurate, sourced, and three of the five are official
The claimMCPs are how you turn Claude from a chatbot into an agent that controls your browser, your repos, your files, and even Blender. Named: the official MCP servers repo, Playwright MCP (Microsoft), GitHub MCP (GitHub official), Blender MCP, Context7.
All five verified real, active, and massively adopted — and three of the five are published by Microsoft, GitHub, and the MCP project itself, which is about as official as receipts get. The core explanation of what MCP is (a protocol giving Claude tool access to external systems) is accurate. The caption's visible keyword-stuffing is an algorithm play, but the links are direct GitHub URLs with no gate and no product. What it skips is only friction: each server needs a config-file edit, a runtime, and 30–60 minutes of setup the carousel doesn't show. Second 'real' in this batch for an MCP listicle (see #072) — the MCP genre is unusually honest, probably because the tools are impressive enough without lying.
What holds up
- modelcontextprotocol/servers, Playwright MCP (~36K stars, Microsoft), GitHub MCP Server (~32K, GitHub official), Blender MCP (~26K), Context7 (~60K) — all real and active
- The capability framing is accurate — this is what MCP actually does
- Direct GitHub links in the post; nothing gated, nothing sold
What doesn't
- Setup friction omitted: JSON config, Node/Python runtimes, and Blender MCP needs a Blender addon too
- Blender MCP is a community project, not official Blender or Anthropic
- Needs a paid Claude plan to use meaningfully
The catch
Only the missing half-hour of setup and the judgment call about access. Start with one server, not five.
Same verdict as #072 and the same receipt: our own daily operation runs on exactly this pattern — the browser automation that fetched this post is an MCP.
- Effort
- 30–60 minutes for your first server
- Cost
- Servers $0 + Claude Pro ~$20/mo
- Stack
- Pick ONE: Playwright MCP (browser) or GitHub MCP (repos) — add more as trust grows
- Confidence
- High
- Posted by
- an AI-prompts account with direct GitHub links and no gate
PARTLY "5 prompts I stole from a friend who's quietly outpacing me" — the prompts are fine; the friend is a writing device
The claim5 prompts stolen from a friend who's quietly outpacing everyone — Brain Dump, Voice Clone, Hard Reply, Lazy Genius workout, Regret Test. Most people don't realize ChatGPT can do this. These aren't basic.
Credit first: the full prompt text is in the caption itself — no 'comment PROMPTS,' no gate. The prompts run fine and two of them (Brain Dump, Hard Reply) are genuinely handy everyday templates. Now the grading: these are standard role-prompting patterns that have circulated in every prompt roundup for two-plus years. 'They're not basic' — they're basic. The 'friend quietly outpacing me' is a stock engagement frame, not a source. And the expert-replacement angle is inflated where it matters: the Regret Test is not a financial advisor and the workout prompt doesn't check your form. Useful starters, oversold origin story.
What holds up
- All five prompts are present in full, ungated — verifiable by reading, and they work as written
- Brain Dump and Hard Reply are legitimately useful daily templates
- Works on free-tier ChatGPT or Claude — the $0 claim holds
What doesn't
- "Not basic" is false — these are the oldest patterns in the prompt-pack genre
- The 'friend' framing is invented social proof
- Expert-replacement framing gets risky at the edges: don't take retirement or medical guidance from a 40-word role prompt
The catch
The only trick here is the packaging: recycled prompts wearing a story. Since the goods are free and functional, the cost to you is just the follow — and believing you've discovered something rare when you've been handed the starter kit.
How to actually do it
- Every 'secret prompt' is three parts: a role, your real context, and an output format. That's it
- Write your own: 'Act as [role]. Here's my situation: [real details]. Give me [format].'
- Feed it your actual writing/numbers/calendar — context is what separates useful from generic, not the prompt's wording
- Where stakes are real (money, health, legal), the prompt is a thinking aid, not the expert
- Anthropic's and OpenAI's free prompting guides teach this in 10 minutes and beat any pack
Skip the mystique — the entire technique in five lines.
- Effort
- 10 minutes to learn the pattern once, then never save a prompt pack again
- Cost
- $0
- Stack
- Free-tier Claude or ChatGPT + your own context
- Confidence
- High
- Posted by
- a prompt-tips account
BS "This prompt jailbreaks Claude to write malware" — the claim we won't help with, and it barely works anyway
The claimA prompt 'jailbreaks Claude [all versions]' into writing code it would normally refuse — shown building a Discord mass-DM/'nuke' bot and a Roblox cheat script.
We grade tactics, not people, and this tactic fails on both honesty and usefulness. We did not capture, test, or reproduce any jailbreak prompt — that's a hard line — so this entry grades the CLAIM, not a recipe. The claim is that one prompt reliably unlocks 'all versions' of Claude to write malware. Its own comment section is the debunk: the top replies point out it only ran on an older model, that newer and Opus-class models don't play along, and that the people who try it get their accounts and game accounts banned. So the pitch — a universal, consequence-free jailbreak — is false on every axis: not universal, not reliable, and definitely not consequence-free. What's real underneath is only the boring, ugly part: some models can sometimes be manipulated, and the AI labs keep closing exactly these holes.
What holds up
- The general phenomenon is real: LLMs can sometimes be manipulated into unintended output — which is why labs run active safety programs against exactly this
- The post's own top comments confirm it's model/version-limited ('he's on the old model for a reason, it doesn't work on the new ones')
What doesn't
- 'All versions' is false by the post's own audience — it's pinned to an older model and fails on current ones
- The output shown (a Discord nuke bot, a game cheat) violates those platforms' rules — commenters note users get banned
- The funnel is the account's other cheat/nuke content and a bio link — the 'jailbreak' is bait for that
- Selling a 'consequence-free' unlock for building abuse tools is dishonest about the actual consequences: bans, and building things that harm other people
The catch
The whole appeal is 'break the rules with no downside.' The downside is right there in the replies — it works on one old model, breaks on the current ones, and gets the people who try it banned. It's a bait post for a channel that sells this stuff, wrapped around a capability that the labs are actively and continuously shutting down.
How to actually do it
- Chasing jailbreaks to build abuse tools gets you banned and builds things that hurt other people — the reels don't show that part
- The models improve specifically against these prompts, so today's 'working' jailbreak is tomorrow's refusal — you're renting a broken tool
- If you're curious about AI safety for real, the honest lane is huge: labs run paid bug-bounty and red-team programs for finding and REPORTING these holes
- Want to build Discord bots or Roblox experiences? The official, unbanned way is the platform's real developer API — and AI helps with that all day
- Judge any 'no consequences' pitch by its own comment section — here, the crowd already called it
What's actually true about AI 'jailbreaks' — and the honest use of that same energy.
- Effort
- N/A — this is a don't
- Cost
- Your account, the target's account, and building tools that harm people
- Stack
- The official developer APIs and, if safety interests you, a lab's red-team program
- Confidence
- High
- Posted by
- an account posting jailbreak and game-cheat content
PARTLY "Stop paying $20/mo for ChatGPT — 5 repos run the model on your machine" — real repos, and the quality trade is the fine print
The claimYou're paying $20/mo for ChatGPT and more for the API. These 5 open-source repos (Jan, Ollama, LocalAI, llamafile, llama.cpp) run the model on your own machine — no subscription, no per-token bill, no account, works on a plane.
The strongest local-AI list we've graded: all five repos are exactly what the caption says (Ollama ~178K stars, llama.cpp ~124K, all pushed the day we checked), the descriptions are honest, and 'start with Jan if you just want the app' is genuinely good guidance. The oversell is the swap itself. A free local model on a normal 8–16GB laptop is a 7–14B-parameter model — genuinely useful, genuinely private, and nowhere near the frontier model your $20 buys. Every serious 2026 comparison lands the same place: local for routine and private work, cloud for the hard stuff. Cancel expecting equality and you'll be disappointed; install expecting a private, offline assistant and you'll be delighted.
What holds up
- All 5 repos verified massive and actively maintained — none archived, all pushed within days
- Every functional description is accurate; Jan's own repo describes itself as 'an open source ChatGPT alternative that runs 100% offline'
- "No account, works on a plane" is literally true
- LocalAI genuinely is an OpenAI-API drop-in for code that expects that API
What doesn't
- The unstated trade: 8–16GB RAM runs small models — the $20 subscription buys a frontier model your laptop cannot host
- Matching subscription quality locally means a 24GB+ GPU box ($1,500–2,500+) plus electricity — 'no bill' becomes 'one big bill'
- Caption funnels to the creator's paid 1:1 calls and app builds
The catch
The dollars move, they don't disappear. You trade $20/mo for either reduced model quality or a four-figure GPU. For private, offline, routine work that trade is genuinely great — the post just never prices it.
How to actually do it
- Install Jan or Ollama tonight — free, genuinely easy, this part of the video is real
- Pull a model that fits your RAM: 8GB = 7B models, 16GB = 12–14B class
- Run it side-by-side against your paid AI on YOUR actual work for one week
- If local covers 80% of your use, downgrade and keep pay-per-use API for the hard 20%
- On 8GB RAM, reframe: it's a privacy tool and a plane tool, not a replacement
The honest way to test the swap before canceling anything.
- Effort
- 20 minutes to first local chat
- Cost
- $0 on hardware you own; $1,500+ if you chase frontier quality locally
- Stack
- Jan (app people) or Ollama (terminal people); llama.cpp under the hood either way
- Confidence
- High
- Posted by
- an AI-content creator (a real engineer) who sells 1:1 consults and app builds
REAL "A DIY smart mirror is the coolest Raspberry Pi project" — it is, and everything claimed checks out
The claimA DIY smart mirror is one of the coolest Raspberry Pi projects — it combines hardware, Linux, programming, and electronics. MagicMirror² displays time, weather, calendar, news, and widgets; easy to customize; great hands-on experience and portfolio material for embedded/software roles.
MagicMirror² is real, free, open source, and the de-facto standard smart-mirror platform, with official docs and a huge module ecosystem. The widget list in the post matches exactly what it does. It genuinely is one of the most popular Pi projects, and it genuinely is legit portfolio material for someone aiming at embedded or software work. The one soft spot: 'easy to customize' is true for the software — the place builds actually stall is the physical half (framing, mounting a monitor behind two-way glass, heat and power management), which the post undersells. No link, no 'comment X,' no funnel. Straight educational content — it happens.
What holds up
- MagicMirror² verified: free, open source, official docs, active module ecosystem
- Feature claims match the software exactly: time, weather, calendar, news, custom widgets
- Cost guides confirm $100–150 for a small build from scratch; $150–400 for a nice ~30" framed build
- Zero monetization in the post — no bait of any kind
What doesn't
- "Easy" covers the software; the woodworking and two-way-glass mounting is where most builds die
- The two-way mirror/acrylic is the wildcard cost item nobody budgets for
The catch
Nothing dishonest — just an incomplete difficulty picture. The software is a weekend; a clean physical build is considerably more, and knowing that going in is the difference between a portfolio piece and a monitor in a closet.
How to actually do it
- Prove the software first: run MagicMirror² on any old monitor or spare tablet — $0 if you own a Pi, and you get the dashboard either way
- Pi 4 (2–4GB) is the sweet spot; a kit runs $50–80
- Only buy the two-way glass after the dashboard's been on your wall for a week and you still love it
- Budget $150–400 for the full framed build and a real weekend-plus for the carpentry
- Put the repo link and a photo on your GitHub — the portfolio claim is real, use it
The build order that avoids the closet.
- Effort
- A weekend for the software; the mirror build is the real project
- Cost
- $100–150 minimum from scratch; $0 to try the software on hardware you own
- Stack
- Raspberry Pi 4 + MagicMirror² + an old monitor first, glass later
- Confidence
- High
- Posted by
- a tech-projects account with no funnel at all
PARTLY "5 AI side hustles that run while you sleep — businesses pay $300–$2,000/mo" — real tools, and the 'sleep' part is the lie
The claim5 AI side hustles that run while you sleep, and the free open-source repo behind each one — Pipecat (voice agents), MoneyPrinterV2 (faceless content), GPT Researcher (research reports), Botpress (chatbots), Skyvern (browser automation). These are jobs businesses already pay $300 to $2,000 a month for.
The $300–$2,000/mo market is real — agencies genuinely charge that for chatbot and voice-agent retainers (we sell in this market ourselves). The repos are mostly real and serious: Pipecat, GPT Researcher, and Skyvern are legitimate maintained frameworks. But nothing here runs while you sleep. Retainers include the parts the video erases: finding the client, onboarding, prompt tuning, and support — that's a job. Worse, one of the five automates a dead strategy: YouTube's July 2025 monetization update explicitly demonetizes mass-produced AI-narrated content, the exact output of MoneyPrinterV2, whose own README says 'educational purposes only.' And Botpress isn't the free self-hosted product implied — the starred repo now feeds Botpress Cloud, a SaaS with paid AI-spend billing. We found zero evidence of anyone earning the promised range passively from these repos.
What holds up
- Pipecat (~14K stars, Daily-maintained), GPT Researcher (~29K), Skyvern (~23K) are real, active frameworks that do what's described
- The service market is real: businesses genuinely pay $300–$2,000/mo for managed chatbot/voice-agent retainers
- The 'free repo behind each paid job' framing is directionally true for three of five
What doesn't
- "Runs while you sleep" describes none of these — client acquisition, setup, and maintenance are the actual job
- MoneyPrinterV2 automates a strategy YouTube policy-killed a year before this was posted; its own README disclaims real use
- Botpress today is a cloud SaaS with usage billing, not the implied free self-hosted tool
- "Free" conceals the running costs: STT/TTS/LLM/telephony APIs — a real voice agent burns $0.05–0.15+ per call-minute
- Zero income receipts — no independent evidence anyone earns the promised range this way
The catch
The pitch sells the retainer market without the retainer work. The $300–$2,000 is real money businesses pay — to people who answer the phone when the bot breaks. That person doesn't sleep through it. Passive income here has an acquisition cost of cold outreach and a maintenance cost of being on the hook.
How to actually do it
- Pick ONE lane where a missed interaction costs a business real money: missed-call answering or website chat for local businesses is the proven entry
- Prototype on vendor free tiers (Botpress Cloud free tier, or Vapi/Retell trial credits) — skip self-hosting Pipecat until you have paying volume
- Land ONE client from your existing network before building anything general
- Price the retainer to cover $30–100/mo of API/telephony costs plus your maintenance hours — that's why it's $300+, not $30
- Skip faceless-content automation entirely — the platform policies got there before you
How the people actually collecting those retainers do it — it's a service business, priced like one.
- Effort
- A working demo in a weekend; the first client is the real work
- Cost
- $30–100/mo to run one real deployment vs the fantasy of $0
- Stack
- One vertical + Vapi/Botpress free tier + a retainer agreement with a human (you) on call
- Confidence
- High
- Posted by
- an AI-content creator (a real engineer) who sells 1:1 consults and app builds
REAL "Cowork didn't replace my work — it took the admin that ate my evenings" — the rare AI take we'd co-sign
The claimClaude Cowork didn't replace my work — it took the admin that ate my evenings. Hot take: the AI fear is aimed at the wrong tasks. It's coming for your busywork, not your judgment.
Same faceless account as entry #068, and this time the framing is right — which is exactly why we grade claims, not people. Cowork's documented sweet spot is precisely folder-shaped admin: file chores, spreadsheet grinding, document drafting, scheduled tasks, with approval gates on anything destructive. Anthropic's own usage data at the web/mobile launch showed most users aren't coding — general knowledge-work delegation is the real pattern. The 'busywork, not judgment' line is fair for today's models and matches what we see running these tools daily. One hedge the post skips: that line has moved every model generation, so don't tattoo it anywhere.
What holds up
- Matches the product's documented capabilities and Anthropic's own usage data
- The evenings-back experience is plausible and mirrors what we see in real client installs
- Approval gates on writes/deletes are real — the human stays on the judgment calls by design
What doesn't
- The busywork/judgment boundary is presented as permanent; it's a snapshot of a moving line
- Testimony from a faceless page is unverifiable — the mechanics check out, the 'my evenings' story can't
- Engagement-bait CTA; no product, no receipts, follower farming is the funnel
The catch
Only the permanence. Today the machine eats formatting and filing while you keep the calls — true. Whether that split holds in two model generations is a bet, not a fact, and the honest version says so.
This claim describes our actual operation: the admin (audits, invoices, briefs, follow-ups) runs on Claude while the judgment calls stay human. We're the receipt.
- Effort
- One folder of real admin work and an afternoon
- Cost
- Claude Pro $20/mo and up
- Stack
- Claude Cowork pointed at the folder where your evenings currently go
- Confidence
- High
- Posted by
- a faceless AI-news page (same account as entry #068 — grading claims, not posters, means this one lands differently)
PARTLY "Start digital marketing for $0 — a 4-step beginner plan" — honest, generic, and quietly not quite free
The claimStart digital marketing for $0 — a simple 4-step beginner plan: pick one niche, make one digital product with Canva + ChatGPT, choose a $0-monthly storefront (Beacons, Payhip, or Gumroad), and post to TikTok to build trust.
This is the mild, honest end of the make-money-online genre — no fake income screenshots, no bundle to buy, no comment-gate, just a reasonable beginner outline. The tools named are all real and the storefronts do have genuine $0-monthly tiers. Two honest caveats keep it from a clean 'real.' 'For $0' skips that those storefronts take a per-sale transaction cut, and Canva/ChatGPT's useful features often sit behind paid tiers. And the plan is generic — 'pick a niche, make a product, post' is true and also the exact advice a thousand identical slideshows give, which is why the hard part (making something people actually want, and getting seen) is the part no 4-step list can hand you.
What holds up
- All tools named are real: Canva, ChatGPT, Beacons, Payhip, Gumroad, TikTok
- The storefronts genuinely have $0-monthly tiers — the 'no monthly fee' claim is true at the base level
- No fabricated income, no bundle, no DM funnel — the honest end of the genre
What doesn't
- '$0' omits per-sale transaction fees on every storefront and the paid tiers of Canva/ChatGPT
- The advice is generic — the same 4 steps every starter slideshow gives; the differentiation you actually need isn't in it
- 'Build trust by posting' is true but silently assumes months of consistency the format compresses to a slide
The catch
Nothing dishonest — just incomplete. The steps are real and free to start; what costs you is the months of making and posting, and the fact that everyone watched the same video. The plan is a fine map; it can't walk the road for you.
How to actually do it
- Pick a niche you actually know something about — that's your only edge over the ten thousand people who saw the same slideshow
- Make ONE product that solves a specific problem for that niche — specific beats broad every time
- Use Gumroad's free tier to start (transaction fee only) — genuinely $0 up front
- Post consistently for real: months, not a week — 'build trust' is the long part the slide hides
- Budget for the fees: every sale gives the platform a cut, and Canva/ChatGPT's best features may want a subscription
The version that gives you a fighting chance, not just a checklist.
- Effort
- An afternoon to launch; months of posting to matter
- Cost
- $0 to start + per-sale fees + optional paid tiers of your tools
- Stack
- Canva/ChatGPT (free tiers) + Gumroad + TikTok + a niche you genuinely know
- Confidence
- High
- Posted by
- a faceless digital-products page building an audience
PARTLY "24 things to install in Claude" — real tools, one genuinely-official catch, and 'install counts' nobody can audit
The claim24 things to install (8 plugins + 8 skills + 8 MCP servers) so one command turns Claude into a full dev, finance, legal, or marketing crew — ship a real workflow by lunch. Named with install counts: gstack (125.5k+), superpowers (264.4k+), codex-plugin-cc ('OpenAI's official Codex plugin'), humanizer (32.4k+), the caveman skill ('65% fewer tokens'), agent-browser (39.6k+), and more.
Most of the tools are real, and the creator earns one genuine point: codex-plugin-cc really is published by OpenAI's own GitHub org — an actual official Codex plugin for Claude Code. But two problems keep this at partly. The install counts are unauditable: Anthropic publishes no official install telemetry, so every '264.4k installs' traces to third-party directories whose method is opaque — superpowers' figure suspiciously matches its GitHub star count. And the caveman skill's headline '65% fewer tokens' is a prose-only number: independent testing measured about 8.5% savings on real agentic work, because code and tool-calls (which dominate) are left untouched. More installs also isn't more capability — every skill loads context every session, and one top comment reports a heavyweight plugin turning a 10-minute task into an 8-hour loop.
What holds up
- Most named tools are real: gstack, superpowers, humanizer, agent-browser, caveman all exist as described
- codex-plugin-cc genuinely IS OpenAI-official — published under the openai GitHub org (the creator got this right)
- 'Turn Claude into a specialist crew' is a fair description of what skills/plugins/MCPs do together
What doesn't
- The install counts are third-party guesses — Anthropic publishes none; superpowers' 264k 'installs' ≈ its GitHub star count
- Caveman's '65% fewer tokens' is prose-only — independent agentic testing found ~8.5%, because code and tool-calls dominate and go untouched
- More plugins = more context loaded every session = more tokens per task, not more capability
- Documented failure mode: heavyweight frameworks front-load thousands of tokens (and can loop for hours) before writing a line
- The real funnel is a comment-'CLAUDE' DM capture, not the list itself
The catch
Installing 24 things is the opposite of the goal. Each one taxes every session's context, and the flashy token-savings and install numbers are marketing, not telemetry. The right move is a handful of tools you understand — the 24-count is the hook.
How to actually do it
- Install 2–3 plugins you'll genuinely use this week — not 24; every SKILL.md eats context on every session
- The one worth knowing about from this list: OpenAI's official codex-plugin-cc, if you also use Codex (note it means paying OpenAI too)
- Ignore the install-count numbers entirely — they're unauditable third-party figures, not usage data
- Test any 'token-saving' skill on YOUR real work before believing the percentage — most savings evaporate on code-heavy tasks
- Watch for runaway loops: cap iterations and keep an eye on the first run of any autonomous plugin
How to actually equip Claude without drowning it.
- Effort
- 20 minutes to install a lean, useful set
- Cost
- Free to install; the cost is token burn from over-installing
- Stack
- Claude Code + 2–3 skills you understand + the official Codex plugin only if you use Codex
- Confidence
- High
- Posted by
- an AI-content creator running a comment-'CLAUDE' DM funnel
REAL "Stop asking Claude for a 10-chapter outline" — a book coach whose free advice is actually the product working
The claim'Give me a 10-chapter outline' is how most people start writing a book with AI — and how they end up with a generic book anyone could have written. Use Claude instead to find your reader's real problem, stress-test each chapter, learn your voice rules first, and cut what readers skip. Claude still can't tell you what readers actually want — that part is your job.
The advice is substantively correct on both craft and mechanics. An unconditioned outline request really does regress to the most average book-shaped output the model has seen — that's how these systems work — and grounding the model in the reader's problem and your voice samples before drafting is exactly the difference between AI-assisted writing and AI slop. The poster even includes the honest limiter most of this genre hides: the model can't tell you what readers want; that judgment stays yours. The funnel is soft (free prompts in bio, surely feeding a list), and the '13 books published' credential is unverifiable — but the value is in the caption itself, ungated, and it stands on its own.
What holds up
- The core mechanism claim is accurate: generic prompt → generic median output is real, documented LLM behavior
- The four recommended jobs (reader problem, chapter stress-test, voice rules, cut what's skipped) match standard nonfiction craft AND good prompting practice
- The honest limiter is in the post itself — rare and creditable
What doesn't
- "13 books published" couldn't be verified — self-pub counts are easy to rack up and impossible to audit from a caption
- Soft lead-magnet funnel (free prompt pack in bio) — fine, but know the free PDF wants your email
The catch
Nothing material. This is what good AI-assisted writing advice looks like: the model drafts, you own the reader and the taste. When a creator hands you the whole method in the caption and the bio link is just the same prompts written down, that's a fair trade.
How to actually do it
- Before anything: paste 2–3 pages of YOUR writing and have Claude extract your voice rules — sentence length, vocabulary, what you never say
- Interrogate the reader first: 'What problem is my reader trying to solve at 11 PM when they grab this book?' — argue with the answers
- Draft chapter by chapter, then stress-test each: 'What would a skeptical reader push back on here? What's missing?'
- Cut pass: 'Mark everything a busy reader would skip.' Believe the marks
- Keep the judgment: the model widens your options; you pick. A book is opinions, and the model doesn't get one
Her four jobs, plus the one setup step that makes them work.
- Effort
- The workflow adds hours; it saves the months a generic draft wastes
- Cost
- Claude Pro $20/mo; the advice itself is free in the caption
- Stack
- Claude + your own voice samples + a reader you can actually name
- Confidence
- High
- Posted by
- a self-published author teaching AI-assisted book writing
REAL "5 of the coolest MCPs let Claude actually DO things" — checks out, and that's worth saying out loud
The claimMCPs are like plugins that let Claude actually DO things — drive your browser, control your files, make 3D, and more. Five of the coolest: Context7, Blender MCP, Playwright MCP, Desktop Commander, Serena — all open-source and free, with thousands more in awesome-mcp-servers.
Every repo exists, every one is major and actively maintained, and every description matches what the tool actually does: Context7 (~61K stars, Upstash), Playwright MCP (~36K, official Microsoft), Blender MCP (~26K), Serena (~28K), Desktop Commander (~9K). MCP itself is the real open standard Anthropic originated, and the 'plugins that let Claude DO things' framing is an accurate plain-English translation. No income bait, no fake numbers. This is what the honest version of the AI-tools listicle looks like — we test enough of the dishonest ones to know the difference. Two footnotes keep it honest: 'free' still requires a paid Claude plan to drive them, and handing an AI your terminal (Desktop Commander) is real power that deserves real caution.
What holds up
- All 5 repos verified real and active within days of this audit — three are backed by Microsoft, Upstash, and established teams
- The capability claims match reality: browser automation, file control, and Blender-by-chat all genuinely work today
- awesome-mcp-servers exists and does list thousands of servers
- We run this exact pattern daily — MCP tools drive the browser that fetched this very TikTok
What doesn't
- "Free" omits the Claude plan (~$20/mo) needed to use any of them
- Desktop Commander gives an AI your terminal — a malicious webpage or document the AI reads can try to steer it; treat it like giving someone your keyboard
- Claude Code and Cowork users already have file + terminal access built in — Desktop Commander is redundant there
- Context7 is open-source but its value is a hosted index — your doc queries hit their API
The catch
None on the facts. The only thing the post can't teach in a caption is the 30–60 minutes of JSON-config setup and the judgment call about how much access to grant. Start with the read-mostly ones (Context7, Playwright) before the ones that hold your keyboard.
Our shop runs on MCPs — browser automation, calendar, email, forms — every single day, including the fetch-and-audit pipeline that produced this entry. The claim isn't hype; it's Tuesday.
- Effort
- 30–60 minutes for a first MCP if you've never edited a config file
- Cost
- MCPs $0 + Claude Pro ~$20/mo
- Stack
- Start with Playwright MCP or Context7; add power tools only as trust grows
- Confidence
- High
- Posted by
- an AI-content creator (a real engineer) who sells 1:1 consults and app builds
BS "One night to decide, hundreds of ready products to resell at 100% profit" — the bundle business where the seller is the business
The claimYou don't need another year of overthinking. Follow 7 steps (pick a niche, faceless IG page, content pillars, daily reels, attach a product, stay consistent) and comment BUNDLE for hundreds of pre-made digital products you can edit, rebrand, and sell as your own for 100% profit.
The 7 steps are generic-but-sound content marketing — nothing false in them, and nothing you can't get free anywhere. They exist to deliver the payload: a master-resell-rights bundle. That model has one reliable earner: the person selling the bundle, who makes money every time someone buys the right to resell what they just bought. The market is drowning in 15M+ identical DFY products, licenses often carry restrictions (minimum prices, no reusing the listing images), and '100% profit' ignores the only number that matters — customer-acquisition cost. Typical outcome for resellers pushing the same files as ten thousand other buyers: roughly nothing, minus ad spend.
What holds up
- The 7 organic-growth steps are directionally real — they're just the same 6–12 months of daily grind every legit creator does
- MRR/PLR bundles are a real product category with real licenses — the mechanics exist
- Outlier success stories exist, which is what keeps the treadmill selling
What doesn't
- The account's own name is 'DFY Digital Products' — the page exists to sell the bundle the post pitches
- Two-stage DM funnel: comment BLUEPRINT (lead magnet) → comment BUNDLE (50%-off tripwire on a made-up anchor price)
- "Sell as your own for 100% profit" is the pitch every buyer up the chain also received — you are the product's customer AND its competition
- Zero revenue receipts offered; 'an audience that trusts you in 6 months' is asserted, not evidenced
- Licenses often restrict pricing and assets — the 'rebrand it' freedom is smaller than sold
The catch
Watch the money's direction: it flows from you to the bundle seller, and your path to getting any back requires out-competing everyone else who bought the same files. The demonstration IS the product — they're showing you the funnel that's being run on you and calling it a business plan.
How to actually do it
- Keep the 7 steps — they're free and they're the real work either way
- Skip the bundle. Make ONE small product that solves a problem you have personally solved for a real person
- Sell it to the audience the 7 steps build — differentiation is the whole game the bundle can't give you
- If you must buy anything, buy back your time: $20/mo of AI to draft your product beats $200 of everyone else's PLR files
- Rule: never buy a business-in-a-box from someone whose business is selling boxes
The version of this that actually compounds.
- Effort
- The same 6–12 months of consistency either way — there is no shortcut in either door
- Cost
- $0 for the honest path vs a $20–500 bundle plus ad spend competing against identical listings
- Stack
- One niche + one real problem you've solved + Claude to draft it + the audience you were building anyway
- Confidence
- High
- Posted by
- a done-for-you digital-products page whose account name IS the funnel
PARTLY "15 tools cost $747/mo — $569 of it is replaceable free" — the list doesn't add up to its own headline, but half the swaps are real
The claimFifteen tools, top-tier priced at $747 a month, but $569 of that can be replaced by free alternatives that do the same job — Notion→Obsidian, Calendly→Cal.com, Dropbox→Nextcloud, Zapier→n8n, Miro→Excalidraw, Otter→Granola, Airtable→Baserow, Typeform→Google Forms, Mixpanel→PostHog, 1Password→Bitwarden, Intercom→Chatwoot.
We did the math the post didn't: the caption promises 15 tools totaling $747, then lists 11 pairs that sum to exactly $569 — four tools and $178 simply never appear. Of the 11 pairs, roughly 9 are genuine and some are excellent (Bitwarden, Cal.com, Excalidraw, PostHog, Google Forms — we run our own scheduling on Cal.com). Two are broken: Granola is neither open source nor free in any real sense (Business is $14/user/mo, the free tier caps your note history), and Mixpanel doesn't even have a $99 tier anymore — it went usage-based with a generous free tier, which deflates its own 'replacement.' Prices also silently mix monthly and annual billing to inflate the total.
What holds up
- 9 of 11 pairs are legitimate swaps; Bitwarden, Cal.com, Excalidraw, PostHog, and Google Forms are genuinely strong free replacements
- Most quoted prices check out at some tier: Notion Business $20, Otter Business $30, Intercom Expert $139
- n8n's free self-hosted Community edition genuinely covers your own commercial automations
What doesn't
- The headline math fails its own list: 11 pairs shown, $569 exactly — the '15 tools / $747' framing is padding
- The Granola pair is flat wrong — it's a paid product with a capped free tier, not a free Otter replacement
- Mixpanel's quoted $99 tier no longer exists — it's usage-based with 1M free events
- Nextcloud, n8n, Baserow, and Chatwoot 'free' means self-hosted: a $5–20/mo VPS plus you becoming your own IT department
- Excalidraw replaces a whiteboard canvas, not Miro's team features; Google Forms replaces data collection, not Typeform's branded UX
The catch
'Free' in these lists means free software, not free operation. The honest saving for a solo operator is real — probably $300–500/mo if you're currently paying top-tier for everything — but you pay for it in hosting dollars and setup weekends, and you were probably never paying top-tier for all fifteen anyway.
How to actually do it
- Do today, zero pain: 1Password→Bitwarden, Calendly→Cal.com, Miro→Excalidraw (for canvas work), Typeform→Google Forms
- Do if you're actually paying for analytics: PostHog's free tier — or check Mixpanel's own free tier first; many teams owe $0 on either
- Transcription: the honest free-ish move is Whisper (open source, local), not Granola
- Self-host n8n/Baserow/Chatwoot only if you enjoy being your own sysadmin — otherwise the subscription IS the cheaper option once your time has a price
- Audit rule: price your CURRENT tier, not the top tier — most 'save $569' math starts by assuming you overpay
The swaps that survive the check, ranked by pain-to-payoff.
- Effort
- An hour for the easy four; a weekend per self-hosted tool
- Cost
- Realistic saving $300–500/mo for a solo top-tier payer; ~$10–25/mo hosting for the self-hosted column
- Stack
- Bitwarden + Cal.com + Excalidraw + Google Forms + PostHog — the five that cost nothing and just work
- Confidence
- High
- Posted by
- a productivity-tools account farming saves
PARTLY "Adobe wants a subscription so you can sign a PDF — these 5 free repos do the whole paperwork job" — great list, false hook
The claimAdobe wants a subscription so you can sign a PDF. These 5 free repos (Stirling PDF — 89K stars, Paperless-ngx, MinerU, Unstract, docling) do the whole paperwork job and none of them send your documents anywhere.
Rare for this genre: every repo is real, huge, maintained, and accurately described — even the 89K star count checked out to the digit (89,385 the day we looked). Stirling-PDF's own README matches the claim almost verbatim, and one Docker command genuinely gets you a self-hosted toolbox covering most 'Adobe jobs.' Two things are still wrong. The hook is false — Adobe's FREE Acrobat Reader signs PDFs, as do macOS Preview and Windows Edge; the subscription was never about signing. And 'none of them send your documents anywhere' breaks on Unstract, whose extraction runs on LLM API calls — your document text goes to the model provider unless you wire up a local model.
What holds up
- All 5 repos verified real and heavily maintained: Stirling-PDF 89,385 stars (claim said 89K — accurate), Paperless-ngx ~44K, MinerU ~77K, docling ~65K (IBM-backed), Unstract ~7K
- Stirling-PDF genuinely covers edit/sign/redact/convert self-hosted — the README quote in the post is accurate
- Paperless-ngx is the real deal for a searchable document archive
What doesn't
- The premise is false: signing a PDF has been free forever — free Acrobat Reader, macOS Preview, Windows Edge, zero install
- The blanket privacy claim breaks on Unstract's default cloud-LLM path
- Three of the five are developer pipeline tools, not paperwork apps — a business owner needs exactly one of these (Stirling)
- Stirling's signing is draw/certificate signing — it does not replace a DocuSign-style audited e-signature workflow with counterparties
- Caption funnels to the creator's paid 1:1 calls and app builds
The catch
This is the 'true list, false frame' pattern: name five real tools, then wrap them in a villain story (Adobe holding your signature hostage) that isn't true. The tools survive the check; the reason you were given for needing them doesn't.
How to actually do it
- Just need to sign something? Free Acrobat Reader, Preview, or Edge — you already own three free signers
- Want a private, self-hosted PDF toolbox? Stirling-PDF via one Docker command covers 90% of it — this part of the post is real
- Drowning in paper? Add Paperless-ngx and give it an afternoon of setup
- Skip MinerU, Unstract, and docling unless you're a developer building document pipelines — and if you use Unstract, know your text is going to an LLM API unless you configure a local model
- Need legally-audited counterparty signatures? That's still DocuSign/Adobe Sign territory — the free stack doesn't cover it
What we'd actually do, by situation.
- Effort
- 15 minutes for Stirling; an afternoon for Paperless
- Cost
- $0 software + Docker comfort + a machine to host on (~$5–10/mo VPS if always-on)
- Stack
- Stirling-PDF alone for most people; Paperless-ngx if you have a real document pile
- Confidence
- High
- Posted by
- an AI-content creator (a real engineer) who sells 1:1 consults and app builds
PARTLY "Claude Cowork has access to your computer and does the work for you" — half right, and the wrong half
The claimClaude Cowork is a Claude that has access to your computer and does the work for you. 9 out of 10 people write it off as a smarter chatbot — it's not.
Cowork is real — Anthropic launched it in January 2026 as 'Claude Code without the code,' built for non-technical people, and it genuinely executes multi-step work: spreadsheets with formulas, documents, browser tasks, scheduled jobs. We run our whole shop on this product family, so we'll vouch for the direction. But 'has access to your computer' gets the key detail backwards: Cowork is folder-scoped and permission-gated by design. It sees the folders you hand it, follows the network rules you set, and asks before anything is permanently deleted. The post erases the exact safety architecture that makes the product handable to a business. And '9 out of 10 people' is an invented stat — nobody measured that.
What holds up
- Claude Cowork is a real Anthropic product (launched Jan 2026), expanded to web and mobile in July 2026
- It does execute real multi-step work autonomously — the 'more than a chatbot' framing is fair
- Included in paid Claude plans from Pro ($20/mo) up — tiers differ by usage, not features
What doesn't
- "Access to your computer" overstates a deliberately sandboxed, folder-scoped design — the safety feature is reported as if it were an omission
- "9 out of 10 people" cites nothing because there is nothing to cite
- Faceless page; the CTA is engagement bait, not information
The catch
The oversell here is subtle: making the product sound MORE powerful than it is also makes it sound less safe than it is. The real pitch — you point it at a folder, it does the work in that folder, and it asks before deleting — is better for an actual business owner than the sci-fi version, and it's the true one.
We didn't have to build the test — this audit itself, the SMS line, the invoices, and the fleet of client installs all run on Claude doing real work on real computers, folder-scoped, every day. The claim's honest version is our daily operation.
- Effort
- Minutes to start: point it at one folder of real admin work
- Cost
- Claude Pro $20/mo and up — not free, and worth saying so
- Stack
- Claude Cowork (or Claude Code for the terminal-comfortable)
- Confidence
- High
- Posted by
- a faceless AI-news page farming engagement
PARTLY "The 4 Levels of Agentic OS 2.0 — Claude Code that runs your work" — a real workflow idea dressed in fabricated dashboard numbers
The claimThe 4 Levels of Agentic OS 2.0 — turn Claude Code from a CLI into a system that runs your work: Skills (workflows that loop and self-improve, 'Claude stops waiting for prompts'), Memory (an Obsidian second brain, cheaper runs), Interface (a dashboard, one click runs skills headless or by voice), Handoff (package the whole OS and share the repo — anyone runs it, no terminal).
The underlying idea is real and good — Claude Code plus skills, a memory vault, and reusable automations genuinely is a step up from typing one prompt at a time, and we run a version of exactly this. Two things pull it down to partly. First, the slides display fabricated dashboard metrics (numbers like '9,875,744' with no source) as if they were live results — that's set dressing, not proof. Second, 'workflows that loop so they self-improve' and 'Claude stops waiting for prompts' oversell autonomy: today's agents run the steps you define, they don't reliably self-improve unattended, and unattended loops are exactly where token bills and runaway behavior come from. The concept ships; the sci-fi autonomy and the invented dashboards don't.
What holds up
- The core stack is real: Claude Code + skills + an Obsidian memory vault + reusable automations is a legitimate, powerful setup
- 'Package it and hand off the repo' is genuinely how these systems get shared with a team
- We run a production version of this exact pattern — the direction is sound
What doesn't
- The slides show fabricated dashboard metrics with no source — decorative numbers presented as results
- 'Workflows that loop so they self-improve' oversells — agents run defined steps; reliable unattended self-improvement isn't here yet
- Unattended loops are where token burn and runaway behavior live — 'Claude stops waiting for prompts' is the risky part, not the feature
- Funnels to a YouTube channel + link-in-bio for 'the full builds'
The catch
The four-level framing is a decent mental model wrapped around invented proof and overstated autonomy. Build the levels — they're real — but verify every automated run, and don't believe the dashboard screenshots any more than you'd believe a stock photo.
How to actually do it
- Level 1 (Skills): codify 2–3 tasks you actually repeat into Claude Code skills — start narrow
- Level 2 (Memory): point Claude at an Obsidian vault with a CLAUDE.md that says how to file and recall notes — this part genuinely compounds
- Level 3 (Interface): a simple script or command to run a skill beats a fancy dashboard; add voice later if ever
- Level 4 (Handoff): document the setup so a teammate can run it — but keep a human approving anything that writes, sends, or spends
- Never let a workflow 'loop to self-improve' unattended — that's the sentence that runs up a bill or breaks something quietly
Build the real version of the four levels, with a human on the loop.
- Effort
- A weekend to a real Levels 1–2 setup; skip the dashboard until you feel the need
- Cost
- Claude plan ($20–200/mo depending on how hard you run it) + free Obsidian
- Stack
- Claude Code + Obsidian + a few verified skills + you on the approve button
- Confidence
- Medium
- Posted by
- a Claude-Code creator funneling to a YouTube + link-in-bio channel
PARTLY "Free repos replace Canva, Placeit, and Kittl for your Printify hustle" — real repos, wrong customer
The claimEveryone is searching how to start a Printify side hustle right now. Printify prints and ships your designs, but the design tools everyone pays for — Canva, Placeit, Kittl — can be replaced by 5 free repos (Medusa, Krita + AI Diffusion, BiRefNet, VTracer, Upscayl), including running your own store without Etsy fees.
All five repos are real, big, and actively maintained — we checked each one the morning we ran this (Upscayl ~48K stars, Medusa ~36K, both pushed within days). Each does roughly what's claimed for its slot. The problem is who can run them. Krita's AI plugin needs a real GPU (6GB VRAM minimum, 12–16GB recommended) or paid cloud tokens; BiRefNet is a Python research model with no app around it; and Medusa is a headless commerce platform that dev agencies build stores on — not something a Canva-tier beginner spins up in a weekend. The stack is pitched at exactly the audience that can't operate it.
What holds up
- All 5 repos exist and are healthy: Medusa ~35.7K stars, Krita AI Diffusion ~10.5K, BiRefNet ~4K, VTracer ~6.6K, Upscayl ~48K — none abandoned
- Upscayl genuinely is a free one-click upscaler anyone can use today
- VTracer genuinely converts raster art to vectors, which print shops do ask for
- No income figure is claimed — that restraint keeps this out of bs territory
What doesn't
- The art-generation step — the whole point — silently requires a gaming GPU (~$300+ used) or ~$11+/mo in cloud tokens
- "Without Etsy fees" hides that Etsy's fee IS the traffic; a self-hosted store starts at zero visitors plus hosting plus Stripe's cut anyway
- Two of five tools (Medusa, BiRefNet) are developer tools with no GUI — the target viewer cannot run them
- AI-generated designs on print-on-demand carry IP-rejection risk nobody mentions
- Caption funnels to the creator's paid 1:1 calls and app builds
The catch
The repo list is honest; the swap isn't. Canva's subscription price is buying you the absence of a GPU, a Docker stack, and a commerce deployment. 'Free' here means 'free if you already own a gaming PC and know your way around PyTorch' — which is not the person searching how to start a Printify side hustle.
How to actually do it
- Keep Etsy/Printify's built-in storefront — the fees are cheaper than building your own traffic from zero
- Use Upscayl (genuinely free and easy) for print-resolution upscales
- Design in Canva Free or Krita WITHOUT the AI plugin; use rembg or Adobe Express's free background remover instead of raw BiRefNet
- Add VTracer only if a printer actually asks for vector files
- Skip self-hosted Medusa until you have sales that justify hiring the developer it requires
The zero-dollar version that a beginner can actually run today.
- Effort
- A weekend to a working shop with the beginner stack
- Cost
- $0 software; the 'free' full stack really costs a $300+ GPU or ~$11+/mo cloud, plus hosting if you self-host the store
- Stack
- Printify + Etsy + Canva Free + Upscayl — boring, free, and runnable
- Confidence
- High
- Posted by
- an AI-content creator (a real engineer) who sells 1:1 consults and app builds
REAL "20 things to tell Claude to add to your website before launching" — a genuinely good pre-launch checklist
The claim20 things to tell Claude to add to your website before launching: custom 404, CTA above the fold, internal links, thank-you page, breadcrumbs, case studies, five FAQs, response-time promise, sticky mobile CTA, robots.txt, unique page titles, meta descriptions, social share image, maps + directions, real reviews, alt text, local schema, privacy policy, Google Analytics, and a real photo of your team.
This one just holds up. All twenty items are legitimate, boring, professional web-launch hygiene — the difference between a site that ranks and converts and one that doesn't. Local schema, alt text, unique titles and meta descriptions, a privacy policy, real reviews, a genuine team photo: this is the checklist a competent web shop runs before going live, and 'tell Claude to add each one' is a real, effective way to get it done. No dollar claim, no product, no funnel — just a fast, correct list. We'll happily co-sign it; we run most of these on every site we build.
What holds up
- All 20 items are standard, legitimate pre-launch best practices — nothing filler, nothing wrong
- Local schema, meta descriptions, alt text, and unique page titles are real SEO fundamentals that measurably help
- Claude genuinely can add every one of these — the workflow claim is true
- No income claim, no product, no comment-gate — rare and clean
What doesn't
- 'In 27 seconds' is the format's pace, not your build time — each item still needs doing and checking
- A checklist tells you WHAT, not whether it's done right — verify each addition actually works (test the 404, confirm analytics fires)
The catch
Essentially none. The only trap is assuming that asking for a thing means you got it — AI will happily tell you it added local schema that has a typo in it. Check the work; the list itself is gold.
How to actually do it
- Paste the 20 items to Claude as a checklist and have it implement them one at a time on your site
- Then VERIFY each: visit a broken URL to see the 404, check the page source for the schema, confirm Google Analytics shows a live hit
- The highest-ROI five for a local business: local schema, real reviews, maps + directions, a real team photo, and one clear CTA above the fold
- Add a real privacy policy with actual content — not a placeholder — especially if you collect any form data
- Don't launch until every item is checked off AND spot-verified — the list is the standard; verification is the discipline
Use it as an actual pre-launch gate — here's how to run it so the items are real, not just requested.
- Effort
- An afternoon to implement and verify all 20 on a small site
- Cost
- $0 beyond your existing Claude plan and hosting
- Stack
- Claude + your site + 20 minutes of clicking to verify the work
- Confidence
- High
- Posted by
- a vibe-coding creator running a checklist series, no product
BS "These tools make running a SaaS worth it" — the only tool identified is the one the post is secretly selling
The claimRunning a SaaS has always been difficult and grueling. However, it is all worth it when you produce results. And these tools allow you to maximize results with minimal effort while keeping your content, you. (Zero tools named in the caption; the slides are headlined by the post's own branded hashtag app.)
We went looking for the headline app and found nothing: no website we could locate, no pricing page, no App Store trail, no reviews — the only place it verifiably exists is its own hashtag. The post format is the standard TikTok seeded-slideshow play: a 'tools that changed my business' carousel where one unknown app rides along with real ones, posted by a faceless account. A tool with no independent footprint can't be evaluated, and a roundup that names zero tools in the caption can't be held to anything. That's not an oversight — the vagueness is the mechanic. You have to swipe, save, and follow to even learn the names.
What holds up
- The pain point is real — running content for a SaaS solo is genuinely grindy, which is why this format farms saves
- Batch-and-repurpose content tools do exist as a category
- The slideshow itself is real and views are real — the product behind it is what we couldn't find
What doesn't
- The promoted app has no findable web presence outside its own hashtag — no site, no pricing, no reviews
- Zero tools named in the caption: the claim is structurally unfalsifiable as posted
- Faceless account + branded hashtag + vague 'these tools' = the textbook astroturf slideshow pattern
- "Minimal effort while keeping your content you" is the exact promise every content-automation funnel makes and none has ever demonstrated
The catch
The post is the product. It looks like a neutral tool roundup, but the only identifiable product in it is the one the account is seeding — and that product can't be checked because it barely exists in public. When a recommendation can't be traced to a real company with a real pricing page, the recommendation is the ad.
How to actually do it
- Record yourself talking about your business on your phone — that's the raw material that sounds like you, because it is you
- Transcribe free with Whisper (open source) or your phone's built-in transcription
- Have Claude or ChatGPT draft posts from the transcript in your documented voice — feed it three examples of how you actually write first
- Schedule free: Buffer's free plan or the native TikTok/Meta schedulers
- Rule for any 'tool list' slideshow: if you can't find the tool's pricing page in one search, it isn't a tool — it's a pitch
The honest version of 'content that stays you' costs zero dollars and uses tools you can actually inspect.
- Effort
- One afternoon to set up a voice-true content pipeline
- Cost
- $0 vs an app with no public price
- Stack
- Phone recorder + Whisper + Claude + a free scheduler
- Confidence
- Medium
- Posted by
- a faceless AI-content page seeding its own app inside a "tools list"
BS "AI agents find 1000s of ready-to-buy leads — 5x humans, 90% cheaper" — four numbers, zero receipts
The claimLet our AI Agents do the prospecting, follow-ups, and booking — automatically. They work 24/7, finding and engaging leads so you don't have to. AI Sales Agents find and engage 1000s of ready-to-buy leads on autopilot — outperforming humans by 5x and saving up to 90% on costs.
The company and the software are real — a contact database plus AI-written cold email at serious mailbox volume, from $59/mo self-serve to $8,000/mo managed. What's graded here are the ad's numbers, and none of them survive. "5x humans" and "90% savings" appear nowhere with a study, baseline, or audit — the site's own case-studies page 404s. Independent data runs the other way: AI SDRs convert meetings to opportunities roughly 40% worse than humans and get flagged as spam at about twice the rate. And "1000s of ready-to-buy leads" describes a scraped contact list with topic-level intent flags. Nobody sells actual ready-to-buy leads for $99 a month — if they existed, they wouldn't need your subscription.
What holds up
- Real company (launched 2023, ~135 staff) — the mechanics of database, AI drafting, automated follow-up, and meeting booking do function
- Some users genuinely book meetings — one verified account ran ~39 meetings/mo at about $31 per meeting
- Strong G2/Capterra scores (4.8/5 across ~600 reviews) — the tool is not vaporware
What doesn't
- "5x," "90%," and a recurring "+312%" carry no methodology, footnote, or third-party verification anywhere
- Category reality: fully autonomous AI SDR tools churn 50–70% of customers a year, and flagship vendors repositioned from "autonomous rep" to "copilot" after the sector's fake-numbers scandal
- The platform's Trustpilot runs much darker than its G2: agencies out $5,000+ reporting zero qualified leads, flagged mailboxes, and refused refunds
- Google/Yahoo bulk-sender rules target exactly this profile — and 500+-recipient cold blasts average 2.1% replies vs 5.8% for small targeted sends; the volume IS the handicap
The catch
The ad sells the removal of the hardest part of B2B — and prices the promise, not the outcome. The tiers are built around 100–250 sending mailboxes, which is a mass cold-email operation with an AI copywriter, not a sales team. Every number in the caption is a marketing number; the numbers that are audited — reply-rate benchmarks, AI-vs-human conversion, churn — all point the other way.
How to actually do it
- Shrink the list, don't grow it: 50–200 hand-verified prospects in a tight ICP. Sub-50-recipient campaigns pull ~5.8% replies; 500+ blasts pull 2.1%
- Use AI for research and drafts only — a human reviews every send and takes every reply. That's the model that survived 2025
- Infrastructure before volume: warmed sending domain, SPF/DKIM/DMARC, one-click unsubscribe, complaints under 0.3% — skip any of it and Google now rejects outright
- Set real expectations: 3–5% replies is normal, 10% is elite; a good month from a tight campaign is a handful of meetings, not "1000s of leads"
- Before paying any AI-SDR vendor: call reference customers, read Trustpilot (not just G2), and never prepay four-figure "lifetime" fees to a vendor with a documented no-refund posture
- Effort
- A tight 100-prospect campaign: 1–2 days to build, 30 min/day to run well
- Cost
- $0–100/mo (email tooling + a warmed domain) vs $59–8,000/mo for the autopilot promise
- Stack
- Any CRM + a cold-email tool with warmup, Claude for research/drafts, a human on replies
- Confidence
- High
- Posted by
- an AI sales-automation platform's own sponsored post
PARTLY TRUE "Claude Code + Higgsfield = €5k website" — the tools are real, the invoice is imaginary
The claimClaude code + Higgsfield = €5k website 🤯. Follow + comment "HOW" and I'll show you how to make a website like this! (over a trucking-company site with a cinematic AI-generated video hero)
Both tools are real and this exact workflow is official: Higgsfield ships Claude Code skills that generate cinematic clips straight from your terminal, and lists website building as a promoted use case. So yes — you can genuinely build the site in the demo. The imaginary part is the price tag. €5k is a fair market rate for a full agency engagement — discovery, conversion copy, SEO, working forms, QA, support. The demo delivers the artifact without the engagement: a landing page with a ~15-second stock-style clip, which maps to the €500–2,500 template tier. And the same pitch circulates across accounts because the platform pays ambassadors up to 15% commission for 12 months per referred subscriber — the trend is the platform's marketing, distributed by people with a cut.
What holds up
- Claude Code really builds and deploys production sites — $20/mo Claude Pro is enough to start
- Higgsfield is a real platform (entry ~$15–19/mo) with an official Claude Code integration — no API key, it burns plan credits
- €3,000–6,000 is a genuine agency range for a small-business build — the anchor is plausible, which is exactly why it works as bait
What doesn't
- The €5k invoice buys discovery, copy, SEO, forms, QA, and support — none of which the one-shot build includes
- Clips cap around 15 seconds, and a usable premium-model clip can effectively cost $50–120 after failed generations
- An autoplay video hero is the single biggest cause of failed Core Web Vitals — the "premium" flourish can cost the client their Google ranking
- "Comment HOW" + an ambassador program paying 15% for 12 months = the poster gets paid on your signup whether or not you ever sell a site
The catch
A trucking company's real €5k site would show their fleet, their drivers, their service area — not a generic AI semi at sunset. The demo looks premium; it doesn't function premium. If you sell these builds at €5k without the agency deliverables behind them, the refund conversation arrives with the first Core Web Vitals report.
How to actually do it
- Replicating the demo costs ~$35–40/mo: Claude Pro ($20) + Higgsfield Starter (~$15–19). The setup is documented free on the platform's own skills page — the "HOW" DM adds a referral link, not information
- Ship the video hero with a poster image and lazy-load, or the clip fails Google's 2.5s LCP line and buries the site
- If you're selling builds: charge for the parts AI doesn't do — discovery, copy that converts, SEO, forms wired to a CRM, browser QA, maintenance. That's what the €5k is for
- If you're buying: pay €500–2,500 for a template-grade site with AI assets, or pay €3–6k and demand the full engagement — just don't pay agency price for the artifact alone
- Spend priority for small-biz conversion: copy first, fast image hero second, cinematic video last — CRO data ranks the flashy part dead last
- Effort
- The demo site: an evening with Claude Code. A site worth €5k: 2–4 weeks including the human parts
- Cost
- ~$35–40/mo for the real stack vs €5,000 of claimed value
- Stack
- Claude Code + Higgsfield skills (official), Netlify/Vercel free tier, a poster-image hero
- Confidence
- High
- Posted by
- a web-dev content page riding a platform-sponsored trend
PARTLY TRUE "FREE Claude-powered plugin edits your videos for you" — real plugin, fake FREE
The claimComment "CLAUDE" for the FREE AutoEdit Content Creator Mode plugin — an AI Premiere Pro plugin powered by Claude AI that automatically edits your videos: cutting bad takes, removing silences, building rough cuts in seconds.
The plugin is real and the tech is legitimate — it wires Claude's API into Premiere Pro for transcript-driven rough cuts, and independent open-source panels prove the same mechanism works. But "FREE" is the funnel bait: the product is a subscription at $9.99–$79.99/mo with monthly video caps, and the free part is a trial. "Powered by Claude" means it calls the API like thousands of other apps — there is no official Anthropic video-editing plugin. And the two loudest promises in the demo, cutting silences and removing filler words, are stock Premiere Pro features you already pay Adobe for.
What holds up
- The product exists and does ship rough cuts, silence/filler removal, and bad-take detection inside Premiere
- Claude-in-Premiere is proven tech: transcript-driven cutting through Premiere's extension API is real, with open-source proof
- Adobe and Anthropic do have a real 2026 integration — but it's a connector in Claude's chat for resizing/cropping social assets, not timeline editing
What doesn't
- "FREE" = a trial on a $9.99–$79.99/mo subscription with caps like 4–40 videos a month
- Not an official Claude product — "powered by" is an API call, not a partnership
- Silence and filler-word removal ship free inside Premiere's Text-Based Editing — the demo's flashiest moments are stock Adobe
- The comment-"CLAUDE" gate is DM automation that captures you as a lead before you ever see a price
The catch
The features the demo leans on hardest are the ones Adobe already gave you. What the plugin genuinely adds — an LLM judging bad takes and taking natural-language instructions — is real but narrower than the reel implies, and it's rentable elsewhere for less. The word FREE is doing all the work in that caption, and it's the one word that isn't true.
How to actually do it
- Already in Premiere, $0: Text-Based Editing → transcribe → filter Pauses + Filler Words → Delete All. That's the silence-cutting demo, native
- Want the Claude layer? Claude Cut (PremiereCopilot) is actually free with a daily quota, $7.99/mo unlimited — Claude-driven bad-take removal with review-before-apply
- Heavier automation: AutoCut ($79/yr, 14-day trial without a card) does silences, captions, take selection, and chapters across Premiere and DaVinci
- If you still want this specific plugin, skip the comment bait and go straight to the product site knowing it's a subscription with a trial — never hand a DM bot your contact info for a link Google finds in five seconds
- Effort
- Native Premiere route: 10 minutes to learn, works today
- Cost
- $0 with your existing Adobe sub; $7.99–79/yr for honest third-party options; $9.99–79.99/mo for the "FREE" one
- Stack
- Premiere Text-Based Editing, or Claude Cut / AutoCut if you want more
- Confidence
- High
- Posted by
- a video-editing plugin studio's promo page
BS "Agents find you on Zillow, $16K in 28 days" — the only real payout is the promoter's commission
The claimIs this the new best side hustle? (GaryVee clip about a kid making $4K cleaning pools) → go to Zillow, agents contact you, dashboard shows estimated revenue $16,408.79 over the next 28 days. Comment "CLAUDE" for the full video. #viewmaxpartner
The tool behind the hashtag is real — a $14–49/mo AI video generator with its own Claude connector and mediocre reviews. Everything else is a mirage. There is no marketplace where Zillow agents find random creators' AI listing tours and pay for them; listing photos belong to photographers and brokerages (Zillow itself is being sued over nearly 47,000 of them), and Zillow's terms prohibit scraping them. The dashboard says "estimated revenue" over the "NEXT 28 days" — a forecast, not a payment, and no third-party account of anyone earning it exists. The one verified money flow in this funnel: the promoter's affiliate program pays 40% lifetime recurring commission on every subscription they recruit. You aren't the partner. You're the recurring revenue.
What holds up
- The AI video tool exists and ships a real Claude integration — the company itself seeds the "Claude side hustle" framing these reels amplify
- Agents genuinely do pay for listing media — when they hire someone they know, with photos they have rights to
- The GaryVee clip is real — a story about a kid doing physical pool-cleaning work for real local customers, the exact opposite of a dashboard paying you
What doesn't
- No ViewMax–Zillow marketplace exists anywhere; the mechanics described are pure invention
- AI tours built from scraped listing photos = unlicensed derivative works — the photo-rights mess is live federal litigation right now
- "Estimated revenue, next 28 days" is a projection screenshot, not a payout; zero independent success reports found
- The affiliate program behind #viewmaxpartner pays the promoter 40% recurring on your subscription — the incentive IS the content
- No GaryVee endorsement of any of this exists; the clip is spliced-in borrowed credibility
The catch
Watch the dashboard wording: real analytics show the LAST 28 days; this one promises the NEXT 28. That single word is the whole scheme — the money is always about to arrive, and the only thing that actually recurs monthly is your subscription charge and the promoter's cut of it. Scam-pattern bonus: agents are actively warning about strangers reposting AI-edited tours of their listings, because rental-deposit fraudsters use exactly that content.
How to actually do it
- The real service exists: local agents pay for listing and social video. But they hire people they've met, using media they have rights to — start with agents in your own network or BNI-style groups
- Get written permission from the listing agent AND confirm photo rights with the photographer before touching any listing media
- Never fabricate interiors with AI on an active listing — label renders as renders. Misrepresenting a property is how you end up in an MLS complaint
- Price as a retainer: 3–5 local agent clients at $300–500/mo is attainable and real. $16K in your first 28 days is not
- If you want AI in the stack, pick the video tool on merit and skip anyone's affiliate link — the tool works the same without paying a stranger 40% of your bill forever
- Effort
- First paying agent client: 2–6 weeks of local outreach and 2–3 portfolio pieces
- Cost
- $14–49/mo for an AI video tool if you want one; $0 to start with a phone and permission
- Stack
- Any AI video tool chosen on merit + written permission + a local agent network
- Confidence
- High
- Posted by
- a side-hustle creator page in an AI video tool's affiliate wave
PARTLY TRUE "My AI salespeople talk to 100s of leads at once, take payments, book calls 24/7" — real software, inflated job title
The claimRight now, my AI salespeople are talking to hundreds of new leads simultaneously. They take payments and book calls on my sales team's calendar 24/7/365.
The core tech is real and commoditized: AI chat/SMS agents that answer a new lead in seconds, qualify, follow up relentlessly, and drop booked calls on a calendar are off-the-shelf products now. What is inflated is the job title. A 2026 study of 828,000 AI-setter conversations across 391 businesses found the agents qualified about 23% of engaged leads and booked roughly 2% into calls — that is a tireless appointment setter, not a sales team. "Takes payments" means the bot sends a checkout link. And texting hundreds of leads sits under a legal framework the ad never mentions.
What holds up
- AI SMS/chat agents that qualify leads and book calls are off-the-shelf tech
- Concurrency is trivially real — software runs hundreds of threads at once
- Speed-to-lead is the genuine edge: first touch in under 10 seconds
- 24/7/365 coverage of inbound response and follow-up is real
What doesn't
- "Salespeople" oversells it — independent data: ~2% of engaged leads booked into calls
- "Takes payments" = sends a Stripe-style checkout link; the AI closes nothing high-ticket alone
- Performs on warm inbound leads; cold-traffic results collapse
- Vendor "15–52% booking rate" figures are marketing numbers on form-fill leads
- The video is an ad for the poster's own AI sales platform
The catch
This is a funnel ad — the poster sells exactly the AI system being demoed, so the "AI sales team" is the product, not a case study. And mass-texting leads in the US runs under TCPA + A2P 10DLC: prior written consent, carrier campaign registration, instant STOP handling, 8am–9pm send windows — with $500–$1,500 exposure per unsolicited text. Unregistered traffic gets blocked by carriers outright. The demo mentions none of it.
How to actually do it
- Stack: GoHighLevel ($97/mo Starter, $297/mo Unlimited) with Conversation AI — or Appointwise layered on GHL — for SMS/DM lead follow-up. Add Vapi or Retell only if you want a voice layer
- Before the first text: register your A2P 10DLC brand + campaign, and only text leads who checked an SMS-consent box on your form. Keep the consent records
- Wire the trigger: form fill or missed call → agent sends first message within 10 seconds. Speed-to-lead is where all the lift lives
- Prompt the agent to qualify on exactly three criteria (budget, timeline, fit), then push the calendar link. "Payments" = a Stripe payment link at the right moment — never card details in chat
- Handoff rule: pricing objection or three unresolved turns → notify a human and stop the bot. This one rule separates working deployments from refund demands
- Set expectations before you build: on warm inbound, ~20% of engaged leads qualify and low single digits book. The economics work on volume and speed, not on a robot closer
- Effort
- 1–2 days to a working setter on warm inbound; 1–2 weeks total incl. A2P registration wait
- Cost
- $97–297/mo GHL + ~$0.01–0.03 per SMS segment; voice adds ~$0.10–0.25/min
- Stack
- GoHighLevel Conversation AI (or Appointwise), Twilio A2P 10DLC, Stripe payment links, Google/GHL Calendar
- Confidence
- High
- Posted by
- founder of an AI sales-automation software company that sells to coaches and course creators
BS "List it tonight, sell 10 a day, $15,000/month" — the math IS the product
The claimBefore you go to sleep tonight, open Claude, create a digital product, use Claude to write the description, and list it on Stan Store for $50. Sell 10 a day. That's $500/day — about $15,000/month. Comment "claude" for the exact prompt.
Every mechanical step is real — Claude can draft a digital product tonight, Stan Store will host it for $29/mo with 0% platform fees. What's missing is the entire business: the reel never mentions where the 10 buyers a day come from. 'Sell 10 a day' is not a step, it's the outcome of having an audience or ad spend the prompt cannot give you. The median outcome of listing an overnight product with no distribution is zero sales, and the comment-gate exists to farm engagement, not to help you.
What holds up
- Claude genuinely can draft an ebook/template/guide + sales description in one evening
- Stan Store: real platform, $29/mo Creator plan, 0% platform transaction fees (processor fees ~2.9% + 30¢ still apply)
- $50 × 10/day × 30 = $15,000 — the arithmetic is fine; the assumption is the fantasy
What doesn't
- Zero mention of traffic, audience, or ads — the only hard part
- The $29/mo subscription and processor fees never appear in the math
- Comment-"claude" gate = engagement farming and a DM pipeline
- Overnight AI products flooding these stores compete with thousands of identical listings
The catch
The reel sells the arithmetic, not the business. 10 sales a day at $50 is what a creator with a real audience earns AFTER years of list-building — quoting the end state as day-one output is the whole deception. If making money were this mechanical, the poster wouldn't need your comment.
How to actually do it
- Flip the order: build the audience first (or borrow one — partnerships, communities you're already in), THEN make the product they ask for
- Use Claude to draft a genuinely useful niche product from your own expertise — specificity is the only moat against the flood of generic AI ebooks
- Validate with 5 pre-sales before building anything — Stan's 14-day trial costs nothing to test
- Budget reality: $29/mo Stan + ~3% processing; break-even is one $50 sale a month, profit is a distribution problem
- Effort
- The product: one evening, genuinely. The 10 sales/day: months-to-years of audience building the reel skips
- Cost
- $29/mo Stan Store + payment processing; $0 for Claude if you already subscribe
- Stack
- Claude + Stan Store (or Gumroad free tier to start at $0/mo + 10% fee)
- Confidence
- High
- Posted by
- a lifestyle side-hustle account
PARTLY TRUE "5 ways to make money with Claude x Higgsfield" — real ways, and the poster is paid by the platform
The claimFive ways to make money with Claude + Higgsfield: animated 3D websites, faceless YouTube channels, UGC ads for brands, product shots for businesses, ad creatives for ecommerce. Comment "guide" for step-by-step start guides. "Of course, these need skill, but skills can be learnt."
The five categories are real, sellable services — SG sells several of them. And credit where due: 'these need skill' is more honesty than most of this genre. What the reel doesn't say: the #higgsfieldcreator tag means the poster is in Higgsfield's Earn program, which pays creators per approved video for posting exactly this kind of content. That's a material connection driving the tool choice — Higgsfield is a paid credits platform, and the 'guide' gate feeds the funnel.
What holds up
- All five service categories are real markets (UGC ads and product shots especially — brands pay $50–500/asset)
- Higgsfield Earn creator program verified: pays per approved video + view bonuses for promotional posts
- The caption's own disclaimer ('these need skill') is accurate and rare for the genre
What doesn't
- #higgsfieldcreator = paid material connection, never stated in plain words
- Higgsfield runs on paid credits — the 'make money' stack costs money monthly
- Comment-"guide" gate: engagement farm + DM pipeline
- Claude's role is real but small here — scripting and prompting; the heavy lifting is the video platform + your reps
The catch
The reel is itself one of the five ways: get paid by an AI platform to post AI content about the AI platform. That's not a scandal — it's how creator programs work — but an undisclosed incentive shaping the tool list is exactly the thing to price in before you buy credits.
How to actually do it
- Pick ONE category with local demand you can verify (product shots for local ecommerce/restaurants is the fastest close)
- Tool-shop on merit: Higgsfield is one option; compare against free/cheaper routes (Claude + open-source video stacks, native Sora/Veo access) before committing to credits
- Portfolio first: 3 spec pieces for real local businesses, then pitch those exact businesses
- Price as a service ($250–500/mo per client), not per asset — recurring beats gig pricing
- Effort
- 2–4 weeks to a sellable portfolio in one category; the 'learn the skill' part is real
- Cost
- Higgsfield credits are paid (plans vary); $0 alternatives exist for most of the five categories
- Stack
- Claude for scripts/prompts + one video/image platform chosen on merit
- Confidence
- High
- Posted by
- an AI-creator account in Higgsfield's paid creator program
PARTLY TRUE "How to create this stunning hero with AI" — a template shop's prompt-gate, the honest-ish kind
The claimThis is how we create stunning hero videos that grab attention. Comment "Prompt" and we'll send the full guide and prompt. Take your website to the next level without it looking like AI slop.
This one's the comment-gate genre (#004, #007, #024) run by an actual business instead of a persona farm — a real Framer template studio with a real marketplace presence, using a free prompt as top-of-funnel for paid templates. The prompt they send is real and will produce a decent hero section. What it won't do is close the gap the ad glosses: one good hero ≠ a site that converts, and 'without AI slop' still depends on the taste of whoever's driving.
What holds up
- The studio is a verified Framer marketplace creator with real templates and components for sale
- AI-generated hero sections from a strong prompt are genuinely achievable — the demo is plausible, not theater
- The funnel is legitimate business development: free prompt → paid template customer
What doesn't
- Comment-gate farms engagement before you get anything
- The prompt is the ad for the template — expect the DM to sell you the $49–99 product
- 'Stunning' is doing heavy lifting: results depend on your eye, not the prompt
The catch
The free prompt makes one section. The paid template is the actual product, and it's a fair product — but the reel frames a storefront as a secret. Same mechanic as the persona-farm prompt-gates we've logged, minus the deception about who's selling.
How to actually do it
- Prompt pattern that works in Claude: describe the business, the ONE action visitors should take, the mood in 3 adjectives, and reference a site you admire — ask for hero copy + layout + motion notes
- For the video-hero look: pair the layout with a subtle background loop (stock or AI-generated), keep text static and readable
- Test it against the SG Site Standard: if the hero doesn't state what the business does in 5 words, redo it
- If you'd rather buy: Framer marketplace templates from real studios are $0–99 and fine — just know that's what the funnel was always selling
- Effort
- 30–60 min to a solid hero with a good prompt; taste is the bottleneck, not the tool
- Cost
- $0 for the prompt route; $0–99 if you buy the template it's advertising
- Stack
- Claude + Framer (or plain HTML/CSS — the prompt pattern is portable)
- Confidence
- High
- Posted by
- a Framer template studio
PARTLY TRUE "If you're selling AI to local businesses, you're a moron" — 50 clients in 21 days, receipts in the paid community
The claimIf you're selling voice agents, chatbots, or websites to local businesses right now, you need to watch this. There's ONE offer that actually closes. Simple test: if a business gives you a dollar, how do you turn it into two? Last 21 days I signed 50+ clients using one offer. Full breakdown inside my community.
The sales principle is genuinely right: local businesses buy outcomes ('a dollar in, two dollars out'), not AI technology — that's how competent agencies actually sell, and the reel diagnoses the tech-first pitch problem accurately. The numbers are the problem: '50+ clients in 21 days' is unverifiable, the account has no footprint outside social platforms, and the only place the 'one offer' exists is behind a community paywall. Signing 2.4 clients a day solo would be a top-1% sales operation — extraordinary claims sold from a hot tub need more than vibes.
What holds up
- The core sales logic (sell ROI, not AI) matches how real agencies close local businesses
- The 'dollar into two' framing is a legitimate qualifying test for any service offer
What doesn't
- '50+ clients in 21 days' — no names, no niche, no revenue, no way to check
- No business footprint findable outside the social accounts
- The entire payoff is gated inside a paid/gated community — the reel is the funnel
- Hot-tub-and-lake set dressing is the aesthetic of income-claim marketing, not of client fulfillment
The catch
The advice is free and correct; the claim attached to it is the unverifiable part. You don't need the community to apply 'sell the outcome' — and if the 50-in-21 number were real, the breakdown would be a case study with named clients, not a paywalled secret.
How to actually do it
- Rebuild your pitch around one measurable outcome (missed calls answered, jobs booked, reviews up) — never lead with 'AI'
- Put a number on the dollar-in-dollar-out: 'the average missed call is worth $X in your trade; this catches them for $Y/mo'
- Proof beats promises: one local case study with real numbers outsells any framework
- If you're evaluating gated communities: ask for three verifiable client names first. Silence is your answer
- Effort
- One afternoon to reframe your offer around an outcome; the discipline to stop pitching tech is the real work
- Cost
- $0 — the correct part of this reel is free
- Stack
- Your existing service + one honest ROI number
- Confidence
- Medium
- Posted by
- a hot-tub-filmed agency guru account
BS "NO CLIENTS? This AI fixes that" — same real tool as #036, now sold with tweet-screenshot theater and a countdown
The claimNO CLIENTS? 6 months building, still $0 revenue. The problem was never the product — it was sales. This AI fixes that: drop your website link, it finds your buyers, writes the emails, books the calls. You just show up & close. $30 free credits, ONLY 9 LEFT — GONE TONIGHT. (Presented via screenshots of viral tweets: 402K views, 'first 3 calls booked in 4 days.')
We already graded the product itself — #036, partly real: a legitimate AI outbound tool with real capability under the hood. This entry grades THIS ad, and this ad is theater. The social proof is screenshots of 'viral tweets' we could not find on X. The scarcity is fake — '$30 free credits, only 9 left, gone tonight' is a standing free-credits offer dressed as a countdown, and the same 'last chance' framing keeps running. When a real product needs manufactured tweets and a fake clock, that tells you what the honest numbers wouldn't.
What holds up
- The underlying product is real and was graded partly real at #036 — capability isn't the issue
- Searched for the screenshot tweets and accounts: no matching viral posts findable on X
- The 'only 9 left tonight' credits offer is the vendor's evergreen trial framing, re-run continuously
What doesn't
- Unverifiable tweet screenshots presented as organic virality — undisclosed ad creative either way
- False scarcity: free trial credits framed as a vanishing countdown
- Preys specifically on the '6 months building, $0 revenue' founder — the most desperate segment
- Cold outbound at '$0.03 per email' still lands in spam-law territory (CAN-SPAM, GDPR) the ad never mentions
The catch
A partly-real product wrapped in a fully-fake ad. Grade them separately: the tool might earn a test (see #036 for the honest breakdown); the marketing tactics on display — invented social proof, fake urgency — are the exact pattern this audit exists to flag. Buy from the pricing page if you must, never from the countdown.
- Effort
- n/a — see #036 for the do-it-yourself outbound stack
- Cost
- n/a
- Stack
- n/a
- Confidence
- High
- Posted by
- the vendor's own ad account
PARTLY TRUE "Claude Video Editor Plug-in" — a real $35 stack, priced honestly, gated behind a Skool funnel
The claimTurn Claude into a 'Godly' video editor with DaVinci Resolve: Claude Code ($20) + Video-Use ($0) + Whisper ($0) + ffmpeg ($0) + Remotion ($0) + HyperFrames ($0) + Higgsfield ($15) + Edits app, Apify, yt-dlp, Resend, Notion (all $0). Comment "AI" for all the resources on my free Skool.
The stack is real and the price list is roughly honest — that's rarer than it should be. video-use (browser-use, 20k+ stars), HyperFrames (HeyGen's HTML-to-video renderer), Remotion's Claude skill, Whisper, ffmpeg, yt-dlp: all real, all free, all genuinely composable into an agent-driven editing pipeline. What's oversold: 'Godly' assembly takes real setup hours per tool, the pieces don't snap together — someone has to wire 12 tools into one workflow — and the comment-"AI" gate feeds a Skool community whose free tier exists to sell you something later.
What holds up
- video-use verified: browser-use's agent video editor, active repo (also graded inside #043/#044)
- HyperFrames verified: HeyGen open-source, ships as a Claude Code skill, single npm install
- Remotion has an official Claude agent-skills route; Whisper/ffmpeg/yt-dlp are the standard free media stack
- Prices check: Claude Code from ~$20/mo, Higgsfield paid tier ~$15 entry — list is materially accurate
What doesn't
- 12 tools ≠ one editor: integration time (a weekend minimum) is the unlisted cost
- 'Godly' output still requires editorial taste — the agent cuts what you tell it to value
- Comment-gate → Skool funnel; free community is the top of someone's paid ladder
- DaVinci Resolve mentioned in caption but does not connect to Claude natively — that part is aspiration, not product
The catch
This is the honest cousin of the fake 'AI edits everything' ads — the tools exist and mostly cost nothing. The gap between 'install 12 tools' and 'have an editing pipeline' is where the Skool community plans to live. You can skip the gate: every tool is public.
How to actually do it
- Start with 3 tools, not 12: Claude Code + video-use + ffmpeg covers 80% (cut silences, pull clips, caption)
- pip install video-use (needs Python 3.11+ and ffmpeg); test on one podcast episode before client work
- Add HyperFrames only when you need motion graphics: it installs as a Claude Code skill via npm
- Ignore the DaVinci tease — export finished cuts and finish color/audio in Resolve manually if you need it
- Effort
- A weekend to a working 3-tool pipeline; the 12-tool 'Godly' version is a multi-week integration project
- Cost
- $20/mo (Claude) + $0 for the core stack; Higgsfield's $15 only if you want AI b-roll
- Stack
- Claude Code + video-use + ffmpeg (+ Whisper for transcripts, HyperFrames for graphics)
- Confidence
- High
- Posted by
- an AI-tools creator running a free Skool community
PARTLY TRUE "Claude Code MCPs everyone should use" — a genuinely good explainer with one sponsored-flavored pick
The claimMCP starter pack — what MCPs are ('AI's plugins that let Claude actually DO things, not just talk') and three everyone should use: Higgsfield (Claude makes the images, videos & ads for your whole site), Playwright (a real browser driven by Claude), Zapier (one plug connects Claude to the 9,000 apps you already run on). Follow for weekly Claude Code tips.
As MCP education, this is better than most paid courses: the 'plugs, but for AI' framing is accurate, and Playwright + Zapier are exactly the two MCPs a beginner should meet first — both real, both free to install, both transformative. The wobble is the third pick: 'Claude makes the images, videos & ads' via Higgsfield overstates it (Claude directs; Higgsfield generates, on paid credits), and Higgsfield's creator program pays accounts for placements exactly like this one. Nothing else is for sale — the only gate is the follow.
What holds up
- The MCP explanation is technically accurate — tools/plugins that let Claude act, not just answer
- Playwright MCP: real, official, free — browser automation from Claude, works today
- Zapier MCP: real, official — connects Claude to thousands of apps (Zapier claims ~8,000; '9,000' is a round-up)
- Higgsfield MCP exists — but generation runs on Higgsfield's paid credits, not your Claude sub
What doesn't
- 'Claude makes the images/videos/ads' — Claude orchestrates; the paid platform makes them
- Higgsfield's creator-payment program makes every organic-looking Higgsfield rec worth discounting
- No setup reality: MCP config still trips beginners (auth, config files) — 'plug it in' is smoother on the slide than the terminal
The catch
Two of the three picks are the right answer and cost nothing. The third is a paid platform wearing the same outfit. That's the pattern to learn from this deck: MCP lists are the new listicle ad unit, and one sponsored pick rides in with two legit ones.
How to actually do it
- Playwright MCP: in Claude Code run claude mcp add playwright -- npx @playwright/mcp@latest — then ask Claude to open a page and click through it
- Zapier MCP: mcp.zapier.com — generate your endpoint, add it to Claude, map your top 3 automations first
- Skip Higgsfield until you have a paying use for AI video — then compare its credits against alternatives on cost per finished asset
- One MCP at a time: each new server adds context overhead; three good ones beat ten idle ones
- Effort
- 15–30 min per MCP including the first successful task
- Cost
- $0 for Playwright and Zapier MCPs; Higgsfield is paid credits
- Stack
- Claude Code (or Claude Desktop) + Playwright MCP + Zapier MCP
- Confidence
- High
- Posted by
- a learn-to-code slideshow account
PARTLY TRUE "Claude + OpenBB = a $25,000 Bloomberg terminal, completely free" — real tool, wrong equivalence
The claimInstall OpenBB, connect it to Claude, ask about any stock, crypto, or filing — Claude pulls live data from 50+ sources and reads it back like a finance analyst. A $25,000 Bloomberg terminal, completely free. Comment "BLOOM" for the link.
OpenBB is real, genuinely open-source, and connecting it to Claude via MCP is a legitimately great research setup — the demo works as shown. The equivalence is the stretch: Bloomberg's ~$25–30k seat buys real-time exchange feeds, fixed-income depth, and the terminal chat network none of which OpenBB replicates. And 'free' has an asterisk — OpenBB's core is free, but many of the '50+ sources' are API keys you bring, several of them paid. It's a free Bloomberg the way a home gym is a free personal trainer.
What holds up
- OpenBB verified: open-source financial platform, actively maintained, MCP integration real
- The workflow shown (ask Claude → query OpenBB → analyst-style readback) works today
- Bloomberg seat pricing ~$25–30k/yr — the anchor number is fair
What doesn't
- Real-time exchange-grade data is NOT free — free tiers are delayed/limited; serious feeds cost money
- '50+ sources' = bring-your-own API keys, several paid (Polygon, FMP, etc.)
- No IB chat, no fixed-income desk tooling, no compliance layer — the parts pros actually pay Bloomberg for
- Comment-gate for a link that's just github.com/OpenBB-finance
The catch
For a solo investor or analyst, OpenBB + Claude genuinely covers the research slice of a terminal — that's a real win worth having. Calling it a Bloomberg replacement is how you disappoint yourself in week two. It replaces the $0-to-$200/mo tier of tools, not the $25k one.
How to actually do it
- Install: pip install openbb (Python 3.10+), or use OpenBB's hosted workspace
- Connect the OpenBB MCP server to Claude Desktop/Code per their docs — then ask for a ticker rundown to smoke-test
- Add free-tier API keys first (yfinance works keyless); pay for a data source only when a specific gap bites
- Reminder that matters here: nothing this produces is investment advice — it's a research tool, and we don't play advisor
- Effort
- 30–60 min to a working Claude-driven research desk
- Cost
- $0 core + your existing Claude sub; optional paid data feeds later
- Stack
- OpenBB + OpenBB MCP + Claude; yfinance/free keys to start
- Confidence
- High
- Posted by
- a startup-team account
HOLDS UP "Anthropic quietly released a plugin that changes Claude Code" — it's real, it's official, it's free
The claimClaude Code feels completely different once you install this. Anthropic quietly released an official plugin called claude-code-setup that scans your project and recommends hooks, skills, MCP servers, subagents, and automations — then sets everything up step-by-step. Most people run Claude Code completely vanilla, which is why it feels messy.
Checked against the source: claude-code-setup is a real plugin in anthropics/claude-plugins-official — Anthropic's own curated plugin directory — and it does what the reel says: analyzes your codebase and recommends the top automations per category (MCP servers, skills, hooks, subagents, slash commands). The install command shown is the real one. Even the editorializing holds: most users DO run vanilla Claude Code, and the ecosystem config is where the leverage lives. No funnel, no gate, nothing for sale.
What holds up
- Repo verified: github.com/anthropics/claude-plugins-official/plugins/claude-code-setup — Anthropic-managed
- Install command verified: /plugin install claude-code-setup@claude-plugins-official
- The official marketplace registers automatically in Claude Code — no extra setup to reach it
- Plugin is read-only analysis + recommendations; it doesn't silently modify your project
What doesn't
- 'Quietly released' is drama — it shipped in the public official directory with docs
- 'Feels completely different' oversells a recommender: it suggests the setup; you still do the configuring
- Recommendations are starting points, not law — blindly installing everything it suggests recreates the mess
The catch
Rare clean one: true claim, official source, zero funnel. The only inflation is emotional ('completely different'). This is what an honest AI-tip looks like — note how little it needed to sell itself.
How to actually do it
- In Claude Code: /plugin install claude-code-setup@claude-plugins-official
- Run it in a real project and ask for its recommendations rundown
- Adopt at most the top 1–2 suggestions per category — the point is fit, not volume
- Revisit after two weeks: keep what you actually used, uninstall the rest
- Effort
- 5 minutes to install; an hour to act on the recommendations that fit
- Cost
- $0 — included with Claude Code
- Stack
- Claude Code + the official plugin marketplace
- Confidence
- High
- Posted by
- an AI-tips account
PARTLY TRUE "$49 beats $4,500 in Google Ads — rank first page FOREVER" — the vendor's math on the vendor's ad
The claimStop wasting money on Google ads. GrandRanker gets better traffic for a fraction of the cost — it creates and publishes SEO content for you, automatically. Set it once. Rank on first page of Google forever. Traffic keeps coming FOREVER, $0 cost per click, you own everything. $49/month vs $4,500/month.
The tool is real — an AI SEO platform that researches keywords, writes articles, and auto-publishes to your CMS, and $49 is a real (promotional, normally $98) price. The physics are the problem: 'set it once, rank forever' describes exactly the autopilot publishing pattern Google's scaled-content-abuse policy targets — sites running high-volume unedited AI content have taken 40–90% traffic losses in enforcement waves, while edited, quality-controlled AI content genuinely works. And '$0 per click' ignores that SEO costs show up as months of lag and content investment before traffic arrives, if it arrives.
What holds up
- Product verified: real platform (keyword research, AI articles, rank tracking, auto-publish to WordPress/Shopify/Webflow)
- $49/mo verified as the current 50%-off promo; regular $98/mo
- The ads-vs-SEO cost asymmetry is a real dynamic — for businesses with content patience
What doesn't
- 'Rank forever' — no one can promise rankings, and 'forever' least of all; Google's spam policy explicitly targets scaled auto-published content
- 'Set it once' is the risk, not the feature: unedited volume is what gets sites hit 40–90%
- Traffic comparison assumes the ads were bad and the content will be good — both unproven
- It's the vendor's own ad; every number is marketing, none are audited
The catch
The honest pitch would be: '$49/mo gets you an AI content pipeline that works IF a human edits it and your niche allows templated content to compete.' That pitch doesn't go viral. 'Forever' does — and 'forever' is the part that gets a site torched in the next spam update.
How to actually do it
- If you try it (30-day money-back): cap output at 2–4 articles/week and edit every one before publish
- Add what autopilot can't: your data, local specifics, real photos, author attribution
- Watch Search Console weekly — declining impressions on new content = stop and reassess, not publish more
- Keep ads for intent you need NOW; SEO is the 6-month lane, not the replacement
- Effort
- Ongoing editorial oversight — the 'set and forget' promise is the thing to refuse
- Cost
- $49/mo promo ($98 regular), 30-day money-back
- Stack
- GrandRanker (or Claude + your CMS doing the same job with more control)
- Confidence
- High
- Posted by
- the vendor's own ad account
HOLDS UP "Claude just outplayed Apple" — true story, verified to the repo. Told by a marketer under an FTC order.
The claimBen Zhang, a robotics engineer, lost his phone at the office. Find My was disabled by his company's MDM. He asked Claude, which suggested tracking the phone's Bluetooth signal strength and built him a live signal meter in about a minute. He walked the office watching the number climb and found the phone. He open-sourced it as 'findphone.'
Every element checks out at the source: the repo is live (ben-z/findphone — 'locate a nearby Bluetooth device by signal strength, from the macOS command line, for when Find My isn't available'), Ben Zhang told the story on his own X account, it hit Hacker News, and the on-screen terminal in the reel matches the actual tool's output (dBm readings, SAME ROOM indicator, peak tracking). MDM disabling Find My is standard corporate practice, and the RSSI approach is sound engineering. The story is true. The teller is the catch.
What holds up
- github.com/ben-z/findphone verified live — macOS CLI, Bluetooth RSSI tracking, exactly as described
- Ben Zhang's first-person account on X + Hacker News thread confirm the origin
- The reel's terminal frames match the real tool's UI (dBm, SAME ROOM, peak/min, ~1s refresh)
- MDM routinely disables Find My on corporate devices — the setup is plausible, not staged
What doesn't
- The poster: Anik Singal settled with the FTC (2023) over Lurn's deceptive income claims — $2.5M redress and orders against misleading earnings marketing (public record: ftc.gov). We name him for the same reason we named the #034 and #035 promoters: federal receipts
- Nothing is sold in this reel — it's authority-farming: true viral stories build the audience the next offer is sold to
- 'In one minute' is roughly right for the first cut, per Zhang himself — but it's Claude + a capable engineer; the minute is not the whole story
The catch
A true story can still be a funnel. This reel is accurate down to the terminal output — and it's posted by a marketer with an FTC settlement on income claims, building AI credibility with other people's engineering. Enjoy the story, verify the storyteller, and remember the two audits where his name already appears.
How to actually do it
- Mac users: clone github.com/ben-z/findphone and run it per the README next time a Bluetooth device goes missing
- The generalizable lesson: when the official path is blocked (MDM, lockouts), describe the physical problem to Claude — signal strength, timestamps, logs — and ask what's measurable
- File under real Claude wins for client conversations: small, verifiable, no income claim attached — the exact opposite of how it's being used in this feed
- Related entries on this poster: #034 (his '$3B AI Berkshire' pitch), and the FTC-receipts standard set at #035
- Effort
- 5 minutes to clone and run if you ever need it
- Cost
- $0 — open source
- Stack
- macOS + findphone CLI (+ Claude for your own version of the trick)
- Confidence
- High
- Posted by
- Anik Singal (Lurn) — named because the receipts are federal record
PARTLY TRUE "AI agents work hundreds of leads at once, 24/7" — true, boring, and sold by a man banned from selling it
The claimAI agents can work hundreds of leads simultaneously, book calls, and send payment links — 24/7, while you sleep.
The capability is real and commodity in 2026 — voice/chat agents that qualify leads, book calls, and send payment links are exactly the stack any competent shop assembles from Vapi/Retell or GHL + Cal.com + Stripe for $100–300/mo. 'Simultaneously' and '24/7' are just what software is. The catch is the seller: the FTC sued his Growth Cave operation (Feb 2025) for ~$50M taken on false income promises, and the January 2026 settlement carries a $48.6M judgment, a permanent ban on marketing or selling business opportunities, and explicit prohibitions on misleading claims about earnings and the use of AI. A true claim about AI from a man banned from making claims about AI isn't information — it's the front end of a funnel.
What holds up
- Multi-lead AI agents with booking + payment links: real, commodity capability (it's literally our Maddox/Bob stack)
- FTC v. Growth Cave/Lee-Tyson: complaint Feb 2025, settlement Jan 2026 — $48.6M judgment, permanent business-opportunity marketing ban (public record: ftc.gov)
What doesn't
- The claim isn't the product — the viewer is; the reel exists to rebuild an audience post-ban
- No mention of TCPA/consent rules, the ceiling every real outbound AI operation lives under
- 'While you sleep' hides the part where somebody builds, tests, and babysits the agent
The catch
Recovered entry — this audit was completed Aug 5 and lost in a deploy gap; the #036 slot it was written for went to another claim, so it lands here with its original date. The verdict aged fine: the tech is real, the pitch is a banned marketer teasing capability he's prohibited from selling as an opportunity.
How to actually do it
- Voice/chat agent: Vapi or Retell (or GoHighLevel's built-in) wired to your calendar — $100–300/mo all-in
- Booking: Cal.com or Calendly link the agent sends after qualifying
- Payment: Stripe payment links the agent can drop in chat/SMS
- The line that keeps you legal: outbound contact needs prior consent (TCPA) — inbound-triggered agents skip the whole minefield
- Effort
- A weekend to a working inbound agent; a couple weeks to trust it with real leads
- Cost
- $100–300/mo depending on stack and volume
- Stack
- Vapi/Retell or GHL + Cal.com + Stripe — the same stack this audit's shop sells and runs
- Confidence
- High
- Posted by
- Lucas Lee-Tyson (Growth Cave) — named because the ban is federal record
PARTLY TRUE "Better than Twilio. Cheaper than Twilio. Ships faster than Twilio."
The claimSent.dm's messaging API beats Twilio on quality, price, and integration speed.
The product is real: one API that takes a phone number, detects whether the person is on WhatsApp, iMessage, RCS, or plain SMS, and routes to the cheapest or richest channel. That's a genuine capability Twilio doesn't offer — Twilio has no iMessage at all. But all three superlatives bend under weight. "Cheaper" holds mainly when messages route to data channels (WhatsApp/iMessage) instead of carrier SMS — for a US business texting US customers over plain SMS, Sent's published ~$0.015-per-contact platform fee plus carrier pass-throughs can land ABOVE Twilio's ~$0.0079/segment. "Better" collapses on reliability: a documented user report describes WhatsApp templates deleted without explanation, chat-only support that never replies, and no status page. "Ships faster" is capped by physics — A2P 10DLC carrier registration takes the time it takes, no matter whose API you call.
What holds up
- Product is real — unified API for SMS, WhatsApp, iMessage, RCS with auto channel detection
- Channel routing to data channels genuinely cuts cost vs carrier SMS, especially internationally
- Twilio's pricing HAS gotten complex: per-message + carrier fees + 10DLC surcharges + number rental
- Twilio offers no iMessage — the channel-breadth gap is real
- Self-serve signup with pay-as-you-go, no sales call required
What doesn't
- "Cheaper" assumes your audience is reachable on data channels — pure US SMS traffic may cost MORE
- Documented user report: WhatsApp templates wiped without explanation mid-production, support unreachable
- No public status page — you fly blind when delivery breaks
- 10DLC registration is a carrier process — no provider makes that wait disappear
- Third-party iMessage sending sits in a gray zone Apple doesn't officially sanction for this category — revocation risk
- Startup vs the market's reliability benchmark: maturity gap is the unpriced cost
The catch
The ad compares its best case (messages routed free over WhatsApp/iMessage) against Twilio's worst case (full SMS carrier fees). Your real bill depends entirely on where your contacts are reachable — and for US-heavy SMS audiences, the savings can invert. The bigger omission: what you pay Twilio for isn't the message, it's the reliability. A messaging provider that deletes production templates and goes silent costs more than any per-message fee.
How to actually do it
- Before touching any provider, profile your audience: pull your contact list and estimate what share is US SMS-only vs WhatsApp/iMessage-reachable. That single ratio decides whether Sent's model saves or costs you money.
- Price your real mix: Twilio US SMS runs ~$0.0079/segment plus carrier fees, 10DLC surcharges, and ~$1.15/mo per number. Sent's published model is ~$0.015/contact platform fee plus carrier pass-throughs. Run both against last month's actual volume.
- If the math favors Sent, pilot with non-critical traffic only — appointment reminders, not payment links. Keep Twilio live in parallel for 60 days.
- Stress-test the failure path before trusting production: file a support ticket on day one and time the response. A provider you can't reach during an outage is not a provider.
- Whatever you pick, the compliance load is yours, not theirs: TCPA consent, opt-out handling, and quiet hours apply identically on every platform. WhatsApp requires pre-approved templates on both.
- For iMessage specifically, know the ground: Apple has no official API for this use case, so every vendor in the category carries revocation risk. Never build a business-critical flow on a channel that can vanish.
- Effort
- 1-2 hrs for the traffic-mix pricing analysis; 1-2 weeks for a parallel pilot
- Cost
- Sent: ~$0.015/contact + carrier fees, pay-as-you-go. Twilio: ~$0.0079/segment + fees + $1.15/mo per number
- Stack
- Your existing contact data + a spreadsheet first; Sent.dm and Twilio side by side only if the math clears
- Confidence
- Medium
- Posted by
- the vendor's own official page — a paid ad from a Twilio competitor
PARTLY TRUE "Pick a task and watch as real work gets done" — save weeks and thousands with Hostinger Agents
The claimHostinger Agents saves you weeks of work and thousands of dollars: pick a task and watch as real work gets done.
The product is real and genuinely cheap: seven specialized AI agents (strategy, content, SEO, marketing, legal, customer comms, sales outreach) for about $7/mo with 1,000 monthly credits. What it actually does is draft — blog posts, keyword lists, privacy policies, emails, plans. You still edit, publish, send, and execute everything. "Real work gets done" sells generation as execution, and "thousands saved" benchmarks against hiring human consultants, not against the general-purpose AI subscription most buyers already have — which matched or beat it in independent head-to-head testing.
What holds up
- Product exists — launched April 2026, live at agents.hostinger.com
- ~$7/mo unlocks all 7 agents, 1,000 credits, no tiered lockouts
- Hands-on testing found outputs usable with only light editing
- SEO agent runs live web search — a real edge over static AI answers
- Auto-detects the right task from plain language, no prompting skill needed
What doesn't
- Outputs are drafts and advice — nothing is executed for you
- Savings math assumes you'd otherwise hire a copywriter, SEO, and lawyer
- Free trial unlocks only 1 of 7 agents with 50 credits
- No money-back guarantee on any plan
- Unused credits reset monthly — no rollover
- Head-to-head test: a raw Claude prompt matched or beat its legal output
The catch
This is the vendor's own ad, and the "thousands of dollars" only materializes if your alternative was hiring three human specialists. The honest comparison is the $20/mo general AI you probably already pay for, where independent testing found comparable output. What $7 actually buys is packaging: auto-triggered skills, live search, and zero prompting required — real value for a non-technical owner, but it's convenience, not new capability.
How to actually do it
- If curious, subscribe MONTHLY only at hostinger.com/ai-agents — $6.99/mo intro, renews $7.99/mo. No refunds exist, so never start on an annual plan.
- Skip the free trial as your evaluation — it unlocks only the Atlas onboarding agent and 50 credits; the six specialists stay locked until you pay.
- Test with real tasks from your own business: one keyword research run, one blog post, one privacy policy, one cold-email draft. Judge by edit-time to publishable, not wow-factor.
- Cancel before renewal if it doesn't beat your current AI — credits reset monthly with no rollover, so there's no reason to bank a second month.
- The $0 alternative: paste the same asks into Claude or ChatGPT with two lines of business context, web search enabled for keyword volume. Independent testing found the outputs comparable.
- Effort
- 10 min to sign up; ~1 hr to test all 7 agents on real tasks
- Cost
- $6.99/mo intro (renews $7.99/mo) — or $0 with the AI you already have
- Stack
- Hostinger Agents (agents.hostinger.com), or any general AI chat with web search
- Confidence
- High
- Posted by
- the vendor's own official page — this is a paid ad, not a creator post
PARTLY TRUE 5 free repos are the entire viral AI video pipeline — all on your own machine
The claimThe talking food, the characters that move exactly how you tell them to — 5 free open-source repos do each piece of the viral AI video pipeline, all of it running locally: DiffSynth-Studio (Wan2.2-Animate motion transfer), EchoMimic (portrait + audio → talking face), video2x (upscaling), LosslessCut (trimming), pyVideoTrans (translate + AI dub).
The repos are real, maintained, and do what the post says — this is one of the more honest tool lists we've graded. What the post skips: the two repos that actually generate video (DiffSynth/Wan2.2-Animate, EchoMimic) need a serious NVIDIA GPU, 50GB+ of model downloads, and a Python/CUDA environment build. 'Free and local' is true only if your machine already has a ~$1,600 graphics card in it. The other three utilities genuinely run on anything. And most of the actual viral talking-food clips are made with hosted models (Veo, Kling, Hailuo), not this stack.
What holds up
- All 5 repos exist, active, genuinely free (Apache/MIT-style licenses)
- Wan2.2-Animate motion transfer is real — ref image + driving video works
- EchoMimicV3 runs on 12–16GB VRAM per its own repo
- DiffSynth has real low-VRAM support: FP8 quantization, layer offload
- video2x, LosslessCut, pyVideoTrans run on ordinary hardware, no GPU wall
What doesn't
- 'On your own machine' assumes an NVIDIA GPU with 12–24GB VRAM
- 50GB+ model downloads before the first frame renders
- Python/CUDA env setup is a real grind — hours, not minutes
- Local render speed and quality trail the hosted tools making the actual viral clips
- No workflow shown — 5 separate tools, you build the glue
The catch
The poster sells 1:1 calls and app builds — the harder this looks to DIY, the more calls get booked. And 'free' quietly excludes the hardware: without a 12GB+ NVIDIA card, your real options are renting a cloud GPU (~$0.35–0.70/hr) or just paying for the hosted tools the viral creators actually use.
How to actually do it
- Check your GPU first: Windows → Task Manager → Performance → GPU. You need NVIDIA with 12GB+ dedicated VRAM (RTX 3060 12GB is the floor, 4090/5090 is comfortable). No NVIDIA or under 12GB → skip to step 5.
- Install ComfyUI Desktop from comfy.org (one installer, handles Python/CUDA for you). Both video models have ComfyUI support — this beats raw-cloning the repos.
- In ComfyUI Manager, install Kijai's WanVideoWrapper for Wan2.2-Animate (motion transfer) and the EchoMimicV3 node pack (talking face). Download the FP8/quantized model variants when offered — they're what makes 12–16GB cards work.
- First test: Wan2.2-Animate wants a clean front-facing reference image (strip the background with rembg or any remover) + a 5–10s driving video at 720p. Expect minutes per clip, not seconds.
- No GPU? Rent one: runpod.io → Deploy → choose RTX 4090 (~$0.35–0.70/hr) → pick a ComfyUI template → same workflow as above, pay only while it runs. A finished 30s clip costs well under $1 of GPU time.
- The three utilities install on any machine, GPU or not: LosslessCut (free .exe from its GitHub releases) for trimming, video2x for upscaling AI output past the 'choppy AI look', pyVideoTrans for translate + dub.
- Honest alternative if this is for client content velocity: hosted gen (Kling/Hailuo/Veo, ~$10–30/mo) makes the same clips with zero setup — the local stack wins on cost only at real volume.
- Effort
- GPU owners: ~half a day to first clip via ComfyUI. No GPU: ~1 hr on RunPod. Raw repos: 1–2 days of env pain
- Cost
- $0 if you own the GPU; ~$0.35–0.70/hr rented; hosted alternative $10–30/mo
- Stack
- ComfyUI + Wan2.2-Animate + EchoMimicV3 + video2x + LosslessCut + pyVideoTrans (NVIDIA 12GB+ or RunPod)
- Confidence
- High
- Posted by
- a prompt-selling creator who books 1:1 calls and app builds off these posts
HOLDS UP “5 AI tools blowing up on GitHub” — checked all five. Real, and bigger than claimed
The claimFive repos are trending on GitHub right now — a full AI agency in your terminal, a free gateway to 231+ AI providers, video editing by coding agent, AI investment research, and Google's own agent CLI — and every screenshot in the post is real.
All five repos verified live against GitHub, and every one is bigger today than the post claimed. agency-agents: claimed 123k stars, actually 139k+. OmniRoute: claimed 9k, actually 43k — and up to 290+ providers since. video-use: claimed 13k, actually 20k. ai-berkshire: claimed 8k, actually 15k. agents-cli is Google's official repo at 5.5k. The ‘REAL SCREENSHOT’ stamps held up. Rare verdict: the post undersold reality.
What holds up
- All five repos exist, active, pushed to this week
- agency-agents: 139k stars vs 123k claimed
- OmniRoute: 43k stars vs 9k claimed — now 290+ providers, 90+ free
- video-use: 20k stars vs 13k claimed
- ai-berkshire: 15k stars vs 8k claimed
- agents-cli: Google's official repo, numbers match
What doesn't
- ‘Whole AI agency’ = specialist subagent prompt bundles, not a business on autopilot
- Caption is recycled save-this bait unrelated to the slides
- Star counts a few weeks stale — in the claim's favor this time
The catch
The inverse catch, for once: the screenshots were real but stale, and every repo has grown since — one nearly 5x. Nothing is for sale here except the follow. When a curation account underclaims and stamps real screenshots that check out, log that too — calibration cuts both ways.
How to actually do it
- agency-agents (specialist Claude Code subagents): git clone https://github.com/msitarzewski/agency-agents, then copy the agent .md files you want into ~/.claude/agents/ (global) or your project's .claude/agents/ folder
- Open Claude Code, run /agents to confirm they loaded, then invoke by name — e.g. ask for the frontend or growth specialist on a real task
- video-use (agent video editing): needs Python 3.11+ and ffmpeg installed; pip install video-use, or clone github.com/browser-use/video-use and follow the README quickstart
- Feed it a source file plus a plain-English edit instruction (‘cut silences, pull the best 45-second clip’) — test on a podcast episode before trusting it on client work
- Skip OmniRoute unless juggling many model providers; skip ai-berkshire unless doing equity research; agents-cli only matters on Google Cloud
- Effort
- 10 min per repo to install; agency-agents useful same-day
- Cost
- $0 for all five repos — you pay only your existing Claude/API usage
- Stack
- Claude Code + git; Python 3.11 + ffmpeg for video-use
- Confidence
- High
- Posted by
- an AI-tools curation account
PARTLY TRUE "Hand off the thinking" — a $67 AI checklist that scales your business for you
The claimAn AI audits your business, then picks from 295 proven systems the exact one to install next — each installs in 5–15 minutes, from a library that's produced $200M+ across 2,500 clients. $67 once, 90-day guarantee.
The mechanic is real and honestly kind of clever. You get a graded library of business systems plus plain-text prompt files you paste into your own ChatGPT or Claude — the AI quizzes you, ranks the library against your business model and revenue stage, and hands you the next system to build. That works. What doesn't hold up: the word "install" and the money math. Pasting an SOP into a doc takes 15 minutes; making a hiring process or a funnel actually run in a real business takes weeks of doing. And the proof number is self-reported and drifts — the ad says $200M+ across 2,500 clients, while the product's own order page says $176M "that I know of." Nobody has audited either figure.
What holds up
- Product is real: graded systems library + AI sequencing prompts
- The "AI" is your own ChatGPT/Claude — an honest, workable mechanic
- $67 one-time with a 90-day guarantee matches the sales page
- Sequencing what to build next is a genuinely useful output
What doesn't
- "Installs in 5–15 minutes" — the document installs; the business change doesn't
- Revenue proof drifts between their own pages ($200M+ vs $176M) and is unaudited
- "Hand off the thinking" — you still make every call and do every build
- Classic $67 tripwire price point; expect upsells behind the checkout
The catch
The AI isn't theirs — it's yours. The $67 buys text files you paste into your own ChatGPT or Claude subscription. The sequencing mechanic is commodity and replicable in an afternoon; the 295-system content library is the actual product, and its "$200M+" pedigree is a number that changes between the brand's own pages and that no one outside the brand can check.
- Effort
- Picker replica: ~1 hr. Any real system: days-to-weeks to actually run, not 15 min
- Cost
- $0 (replica) or $67 one-time (theirs) — plus the ChatGPT/Claude sub you already pay for
- Stack
- Plain-text prompt files + your own ChatGPT or Claude. Nothing proprietary
- Confidence
- High
- Posted by
- a business-systems brand running paid ads for a $67 checklist product
PARTLY TRUE "A full 95-minute AI feature film just got made" — real film, inflated folklore
The claimA complete 95-minute AI-generated feature film was just made and shown at Cannes — proof AI can now produce full movies.
The film is real; the folklore around it isn't. 'Hell Grind' exists — a 95-minute AI-generated sci-fi feature built by a 15-person team at an AI video company using a next-gen video model, shown in Cannes in May 2026. That's a genuine milestone: character consistency and a narrative arc sustained across feature length, which was the field's hardest bottleneck a year ago. But every viral retelling inflates three things. It didn't 'premiere at the Cannes Film Festival' — the festival publicly confirmed it was never in the official program; it screened at a rented commercial cinema in town during the market. It wasn't push-button: prompts averaged ~3,000 words of cinematography spec per 15 SECONDS of footage, with a human director and co-writer with real festival credits steering it. And it wasn't free: roughly $500,000, about $400k of it pure compute, over 14 days. The honest headline is 'AI just made feature films radically cheaper and faster for skilled teams' — not 'anyone can prompt a movie.'
What holds up
- 95-minute AI feature genuinely exists and screened in Cannes (the town)
- 15 people, 14 days — a real order-of-magnitude compression vs. traditional production
- Feature-length character/world consistency is a legitimate technical first
What doesn't
- Cannes festival publicly denied it was in the official program
- ~$500,000 cost (≈$400k compute) — not a hobbyist budget
- ~3,000-word prompts per 15-second shot; professional director and co-writer required
- Poster monetizes AI-authority content — the milestone doubles as funnel fuel
The catch
The demo cut away from the invoice. Half a million dollars, 80% of it compute, plus fifteen working filmmakers writing novella-length prompts for every fifteen seconds of screen time. The capability leap is real — but the moral being sold ('AI makes movies now, so buy my AI thing') skips that the scarce ingredient is still taste, direction, and a compute budget.
How to actually do it
- Ignore feature-length — the commercial sweet spot right now is 15–60 second spots, where these same models (Higgsfield/Seedance-class, Veo, Kling) are production-ready.
- Treat prompts like shot lists, not wishes: camera, lens, motion, lighting, wardrobe continuity. The 3,000-word-per-shot detail IS the lesson — specificity is the skill.
- Budget compute honestly: expect $50–300 in generation credits per polished 30-second spot after retries, not $5.
- Keep a human editor in the chain — the Cannes team shipped because filmmakers curated thousands of takes, not because the model got it right once.
- Effort
- A polished 30s AI spot: 1–2 days for a practiced operator
- Cost
- $50–300 per finished spot in model credits; tool subscriptions $10–100/mo
- Stack
- Higgsfield / Veo / Kling class video models + a human editor with taste
- Confidence
- High
- Posted by
- a longtime internet-marketing figure turned AI investor (large following)
PARTLY TRUE "$2,500+/mo clients with ZERO fulfillment" — the data-as-a-service agency pitch
The claimOver 1,000 agencies have white-labeled 'Data as a Service' to land clients paying $2,500–$10,000+/mo every week — with zero fulfillment and no overhead. Apply to see if your agency qualifies.
The model exists; the math and the 'zero' don't. Identity-resolution platforms are real: a pixel matches anonymous website visitors to name/email/address records, and agencies genuinely resell that data access as a recurring line item. But this reel is the vendor advertising its own platform — the 'apply to qualify' funnel leads to demo-gated, undisclosed pricing that third-party reviews peg in the four-to-five-figure range, with no published refund policy. 'Zero fulfillment' is false on its face: you still prospect, close, build audiences, sync them, and consult — the vendor's own case studies describe account managers and ad spend as ongoing costs. And '1,000 agencies at $2,500+/mo' is a vendor-sourced number with no outside verification. The biggest omission is legal: reselling resolved identities of people who never opted in sits in a privacy-law gray zone the reel never mentions.
What holds up
- Visitor identity resolution is a real product category
- Agencies do resell data access as recurring revenue
- 'Sell data, not hours' genuinely reduces delivery labor vs. full-service
What doesn't
- Vendor's own ad graded as if it were a case study
- Pricing hidden behind a demo; reviews report $10k+ spends and no refunds
- 'Zero fulfillment' contradicted by the vendor's own case studies
- '1,000 agencies / $2,500+/mo' unverifiable, vendor-sourced
- Person-level visitor data triggers CCPA/state-privacy and CAN-SPAM/TCPA exposure on cold outreach
The catch
This isn't an agency success story — it's the platform's ad for itself. The 'zero-fulfillment agency' is the bait; a demo-gated, high-ticket subscription is the product. And the line item nobody prices in: emailing or calling people whose identities you resolved without consent is where privacy regulators live. The reel sells the revenue and skips the liability.
How to actually do it
- Start company-level, not person-level: tools like Leadfeeder or Vector identify which COMPANIES visit a site — clearly lawful B2B intel, no individual identities. Pilot on your own site first.
- Package it as a $500–$1,500/mo add-on for existing clients already spending on ads: 'here's who visited and didn't convert' + a monthly action call.
- If you ever go person-level (RB2B-class, US-only), have a privacy attorney bless your disclosure, opt-out, and suppression flow BEFORE any outreach touches that data.
- Call the offer what it is: light-fulfillment productized service. Audience build + sync + one consult per month is the real scope — price that honestly and it still carries 70%+ margins.
- Effort
- 1–2 weeks to pilot company-level on your own site; 30–60 days to first client add-on
- Cost
- $0–200/mo to start (company-level tools have free tiers); the advertised platform: undisclosed, demo-gated
- Stack
- Leadfeeder or Vector (company-level ID), existing client ad accounts, a privacy attorney before anything person-level
- Confidence
- High
- Posted by
- a data platform's own recruitment ad (vendor-run, application funnel)
HOLDS UP "Claude + Granola MCP cheat code" — a paid ad that's actually telling the truth
The claimGranola is an AI notepad that takes great meeting notes WITHOUT sending a bot into your meetings — and pairing it with Claude over MCP is a 'cheat code': Claude can pull all of yesterday's meetings and turn them into scheduled tasks and briefs.
Rare species: a sponsored post where every functional claim checks out. Granola runs as a desktop app that transcribes your computer's audio directly — no bot joins the call, no participant-list entry, which is its core documented differentiator, not ad gloss. And the 'cheat code' is just Granola's own official MCP server (launched Feb 4, 2026, mcp.granola.ai/mcp) doing exactly what the video shows: Claude lists yesterday's meetings, pulls notes and action items, drafts the follow-ups. We run the same class of setup daily. What the ad skips: the full MCP experience sits on the Business plan at $14/user/month — the free tier caps MCP at your own last 30 days of notes with no transcripts — and 'no bot in the meeting' cuts both ways: nobody on the call knows they're being transcribed. Roughly a dozen US states require all-party consent for recording conversations. Granola itself concedes you 'may want to inform participants' and offers an optional extension that announces it. Optional is the operative word.
What holds up
- Granola genuinely captures meetings without a bot — system-audio transcription, no participant-list entry (its own docs)
- The official Granola MCP server is real (Feb 2026) and works from Claude, Claude Code, and ChatGPT — the demoed workflow is the product working as designed
- It's a properly DISCLOSED paid partnership with a link card — more than most AI ads manage
- The company is legit: $125M Series C at a $1.5B valuation, March 2026
What doesn't
- Full MCP access is a paid feature: Business plan $14/user/mo; free tier = 30 days of personal notes, no transcripts
- Bot-less capture means call participants get NO indicator they're being transcribed — an all-party-consent problem in states like CA, FL, WA, IL, PA
- Audio is processed by a cloud transcription provider — it is not local-only, whatever 'notepad' implies
- 'Cheat code' is influencer-speak for 'documented official feature'
The catch
The selling point and the ethical trap are the same feature. 'Doesn't join your meetings' means it also doesn't announce itself to the people in them — in an all-party-consent state, silently transcribing a client call is legal exposure, not a productivity hack.
How to actually do it
- Say 'I'm running an AI notepad on this call' at the top of every meeting. Five seconds, and it converts the legal risk to a professionalism flex
- Budget the $14/user/mo Business seat if you want the full Claude integration — the free tier's 30-day window will feel broken the first time you need last quarter's call
- Setup is one line: add the 'Granola' connector in Claude settings, or claude mcp add granola --transport http https://mcp.granola.ai/mcp
- Test the recall before you trust it: ask Claude for yesterday's action items and check them against your own memory of the call once
This one you can actually copy — just do it with consent and a budget line.
- Effort
- 15 minutes to install and connect; the habit of announcing it costs nothing
- Cost
- Free tier to trial; $14/user/mo Business for the real MCP workflow
- Stack
- Granola desktop app + official Granola MCP + Claude (any surface); optional notification extension for meets
- Confidence
- High
- Posted by
- a small self-improvement podcast page (~700 likes) running a disclosed paid partnership with Granola
PARTLY TRUE "The White House ordered AI into every small business" — no, it didn't. And the 'model' pays him twice.
The claimThe White House released an official AI Action Plan directing federal agencies to drive AI into U.S. small businesses. 92% of the 36 million small businesses have no AI strategy; only 3% of consultants offer implementation. Someone gets paid $2,500/mo to close that gap — the model: find businesses with Claude + Google Maps, close a 15-minute call, have a $6/hr OnlineJobs.ph contractor install a pre-built stack in 14 days, charge $1,500–$3,000/mo recurring.
The document is real. The directive is not. 'Winning the Race: America's AI Action Plan' exists (July 2025, ~90 federal actions) — we read all 28 pages. The ONLY small-business provision in it is a line telling Commerce/NTIA to 'convene stakeholders to help drive adoption of open-source and open-weight models by small and medium-sized businesses.' One convening, about open-weight models. No mandate, no program, no money pointed at Main Street installs — the plan literally names LARGE organizations as the adoption problem. His stats dissolve on contact: no survey anywhere says '92% have no AI strategy' (the likely source is Census data showing 8.8% of small firms USING AI — inverted and relabeled; the US Chamber has 40% of SMBs using gen AI). '3% of consultants offer implementation' has no source in existence. Credit where due: '36 million small businesses' is the current SBA count (36.2M, June 2025) — though ~27M of those are solo side-gigs, not shops with $2,500/mo to spend. Then the model: your client hands admin access to their CRM, email, and customer data to a $6/hr overseas contractor who's never seen the business, while the poster collects twice — an undisclosed-price coaching program closed on a 'strategy call,' plus OnlineJobs.ph's 40% RECURRING affiliate commission on the link he embeds in the funnel.
What holds up
- The White House AI Action Plan is real — July 23, 2025, three pillars, 90+ federal actions
- 36 million small businesses is the current SBA figure (36.2M as of June 2025) — he's more current than most skeptics
- The gap is genuinely real: Census says under 9% of small firms actually use AI in operations, and most that do have no formal strategy
- AI-implementation-as-a-service is a legitimate business — it's ours
What doesn't
- 'Directs agencies to drive AI into small businesses' = spin on ONE open-source convening provision; the plan's own text says the adoption worry is big orgs, not Main Street
- '92% no AI strategy' and '3% of consultants' — no source exists for either; the precision is costume
- The OnlineJobs.ph link card pays him a 40% recurring commission whether or not you ever buy his program
- His program has no public price — application, then a 1-on-1 'strategy session' close, aimed at '$150k+ corporate professionals'
- Same person previously sold a 'Modern SDR' course and a 'remote triaging' opportunity — new wave, same funnel
- The delivery model hands a US small business's customer data to a $6/hr offshore contractor with zero local accountability — that's where the widely-cited '95% of AI pilots fail' stories come from
The catch
Follow the money in both directions: you pay an undisclosed high-ticket fee to learn to charge a business $2,500/mo, a $6/hr stranger does the install, and he collects 40% recurring on the VA site he just taught you to need. The White House appears in this post for one job: making all of that feel inevitable.
How to actually do it
- Before any 'the government just changed everything' post: pull the actual document. The plan is a free 28-page PDF on whitehouse.gov — the small-business section takes 30 seconds to not find
- When a stat sounds load-bearing ('92%', '3%'), search it with the word 'survey'. If the only source is the post itself, it's decor
- Check what the poster earns from the links: OnlineJobs.ph pays 40% recurring to affiliates — an embedded link card is a revenue channel, not a resource
- The legit version: scope the business's actual bottleneck first, install and test personally, keep the data inside the business, document everything, train the owner, and price on delivered hours saved — not on a stat nobody can source
The honest version of this business exists. It's boring in the best way, and it can't be run from a beach.
- Effort
- His version: 14 days of somebody else's labor. The real version: weeks of on-site scoping, install, testing, and training per client — that's why it's worth paying for
- Cost
- His program: undisclosed, call-close only. The honest alternative: a real local install runs $1,500–$4,500 + monthly care, priced on outcomes
- Stack
- What he's selling: Claude + Google Maps prospecting, a 'pre-built stack,' OnlineJobs.ph contractors ($6/hr), a $300/mo VA — and a funnel where the affiliate link is the most reliable revenue line
- Confidence
- High
- Posted by
- a ~900K-follower 'AI implementation' coach (verified) running a classic comment-funnel with an affiliate link card
HOLDS UP "6 repos I'd install before paying for another subscription" — and for once, all six are real
The claimSix open-source GitHub projects can replace paid subscriptions: local voice cloning/dictation (Voicebox), internet research for AI agents (Agent-Reach), anti-slop web design (Hallmark), job-search automation (career-ops), agent-built video (HyperFrames), and a Bloomberg-style intelligence dashboard (World Monitor).
Rare one: every repo checks out. All six exist on GitHub, all are actively maintained (pushed within the last two days), all are MIT/Apache or similar, and every star count on the slides is accurate or now higher — Voicebox 49k, Agent-Reach 67k, Hallmark 22k, career-ops 63k, HyperFrames 40k, World Monitor 79k. The poster even discloses that some take real setup. The overselling is in the word "replace": local voice cloning needs capable hardware and still trails ElevenLabs on quality; career-ops runs inside an AI coding CLI like Claude Code — which is itself a subscription; World Monitor is a news/OSINT dashboard, not a Bloomberg Terminal (no market data feeds, no execution); and "free" open source still means your own compute or API token costs. Free as in puppy, not free as in beer.
What holds up
- All 6 repos verified directly on GitHub — real, public, active
- Star counts on slides accurate; most have grown since posting
- All permissively licensed (MIT / Apache 2.0)
- Poster discloses setup effort — no get-rich framing
- Repo links given away free, no email gate, no visible course funnel
- HyperFrames is HeyGen's own open-source project — corporate-backed, not abandonware risk
What doesn't
- "Replace ElevenLabs" oversold — local cloning needs a real GPU or Apple Silicon and trails on quality
- career-ops requires an AI coding CLI subscription to run — replacing one subscription with another
- World Monitor is not Bloomberg — no market data terminal, no trading; the $32k comparison is theater
- "Free" excludes compute, API tokens, and your setup hours
- Six installs is a hobby; most people need one or two of these
The catch
Nothing sinister — no course, no funnel. The only spin is the word "replace." These are excellent free alternatives at 80% of the paid product, and the missing 20% is exactly the part you were paying for: polish, hosting, and zero setup. Budget an afternoon per tool, not five minutes.
How to actually do it
- Agent-Reach (research agent superpowers): needs Node 18+. Run `npx agent-reach install` in a terminal, or add it as an MCP server to Claude Code — repo README at github.com/Panniantong/Agent-Reach has the one-line MCP config. Zero API fees; it scrapes directly, so respect platform ToS for commercial use.
- Hallmark (anti-slop design checks): in Claude Code run `/install github.com/Nutlope/hallmark` or copy the skill folder into ~/.claude/skills. It then auto-runs 57 design checks on any site your agent builds. Free, no accounts.
- HyperFrames (HTML-to-video): `npm install hyperframes` in a project, write scenes as HTML/CSS, render to MP4 locally. Docs at github.com/heygen-com/hyperframes. Needs Node + ffmpeg installed.
- Voicebox (voice cloning/dictation): desktop app download from voicebox.sh — needs Apple Silicon Mac or a Windows machine with an NVIDIA GPU for usable speed. Voice cloning laws apply: clone only voices you have rights to.
- career-ops and World Monitor: personal-use tools. career-ops installs into Claude Code (`/install github.com/santifer/career-ops`); World Monitor is `git clone` + `npm run dev`, or just use the hosted version at worldmonitor.app free.
Verified all six against GitHub primary sources. Flagged the three with real business value for a consulting shop — Agent-Reach (agent research without per-call API fees), Hallmark (design-quality checks on AI-built client sites), HyperFrames (agent-generated video for content) — with exact install steps. The other three are personal-use tools.
- Effort
- 20-60 min per tool to install; an afternoon each to make useful
- Cost
- $0 software; your compute + existing Claude Code sub for two of them
- Stack
- Node 18+, Claude Code (for Hallmark/career-ops), ffmpeg (HyperFrames), GPU or Apple Silicon (Voicebox)
- Confidence
- High
- Posted by
- a faceless AI-tools slideshow account
PARTLY TRUE Explee's AutoGTM — "a 24/7 AI agent that finds clients while you sleep" for $0.03 an email
The claimPaste your website URL and a team of AI agents researches your market, finds high-intent prospects, writes personalized cold emails, handles follow-ups and domain warming, and books demos on autopilot — pay-as-you-go at about $0.03 per email, no subscription.
The product is real and the mechanics mostly check out — Explee LTD is a registered UK company with a live pricing page, an API, and a 105M-company / 536M-profile database behind a genuinely useful semantic search. The hype is in the outcome framing. Their own pricing page does the debunking for them: $30 buys 1,000 emails, which they project at 2–8 warm leads and 1–2 meetings. That's a real pipeline trickle, not 'clients while you sleep' — and whether even that materializes depends entirely on your offer and targeting, which no agent fixes for you.
What holds up
- Real registered company (Explee LTD, London) with public pricing and API docs
- AI Search over the company database works and has a free tier — no signup required
- Pay-as-you-go pricing is real: ~$0.03/email, $50 free credits for new accounts
- Pre-warmed mailboxes and automated follow-ups are a real, shipping feature
- Email verification via ZeroBounce; sub-3% bounce claim is plausible for verified data
What doesn't
- Their own math: 1,000 emails → 1–2 meetings. A ~0.1% meeting rate is the honest number under the hype
- 97% deliverability is the vendor's own figure — no third-party audit
- All glowing testimonials live on the vendor's homepage; only a handful of G2 reviews exist
- Cold email at volume has a legal ceiling (CAN-SPAM, GDPR for EU targets) — and their terms of use push all compliance liability onto you
- Sending from vendor-managed mailboxes means your brand reputation rides on their infrastructure
The catch
The ad sells the dream of autonomous client acquisition, but the vendor's own pricing page quotes 1–2 meetings per 1,000 emails sent. The agent automates the sending — it does not automate having an offer people want. And every compliance risk of mass cold email is contractually yours, not theirs.
How to actually do it
- Free win first: go to explee.com/landing-search (no signup needed) and run a plain-English ICP search — e.g. 'multi-truck home service companies in Northern Colorado with 10+ employees'. Judge data quality on names you already know.
- If the data holds up, create a free account: 500 welcome credits, no credit card.
- For the AutoGTM test: sign up at explee.com, use the $50 new-account credit or put in $30 — that's ~1,000 emails.
- Point it at ONE specific offer with a clear call to action, not your whole business. US-only targeting to stay out of GDPR territory.
- Set the daily budget cap low (it's in the campaign settings) and read the first batch of AI-written emails before letting it run — your name is on them.
- Success bar from their own numbers: 1–2 booked meetings per $30. If a test booking costs you under $15, it beats most paid channels; if it returns zero after 1,000 sends, kill it — the data was fine, the offer wasn't.
Worth a controlled test, not a leap of faith. The free AI Search tier is the no-risk win — use it for list-building today. AutoGTM itself gets a capped $30–50 experiment on one tight offer before any real budget.
- Effort
- 10 min for the free search test; ~1 hour to set up and review an AutoGTM campaign
- Cost
- $0 to evaluate; $30–50 for a real test; ~$30 per 1,000 emails ongoing
- Stack
- Explee AI Search (free tier) + AutoGTM pay-as-you-go; your own calendar link for bookings
- Confidence
- High
- Posted by
- the vendor's own Facebook ad account
BS "I cancelled $211/month in AI subscriptions" — the $8 replacement sold by its own CEO
The claimI cancelled $211/month in AI subscriptions (ChatGPT Plus, Midjourney, ElevenLabs, Suno, a video AI). One tool replaced all of them: Supercool, $8/month. Music, video, voiceover, photos, code, scripts, a full LLM — all in one workflow. No AI tool beats Supercool. Not Claude. Not GPT. Nothing.
The tool exists. Everything load-bearing about the post is deceptive. Start with who's talking: the 'I cancelled my subscriptions' testimonial is written by Alex Mehr — the CEO of Famous Labs, the company that owns Supercool. That's not a review, it's an undisclosed native ad by the owner, which the FTC's endorsement rules call deceptive on its face. The same promoter was charged by the SEC in September 2025 (alongside Tai Lopez) over the RadioShack/Pier 1 revival — ~$112M raised from investors, 'Ponzi-like payments' alleged. Then the price: $8 doesn't buy the product, it buys starter CREDITS — and auto top-up is switched on by default. Trustpilot is stacked with the receipts: one user charged $534 in about 20 minutes processing a single document; another auto-billed $600+ into overdraft. The economics make the headline impossible — state-of-the-art video generation alone costs more per minute in raw API fees than $8. 'Beats Claude and GPT' from a tool that wraps somebody else's models is pure puffery — it won't even say whose models it wraps.
What holds up
- Supercool is a real product that does generate music, video, voice, images, and text
- The itemized subscription prices in the post are roughly real market prices
- The company answers Trustpilot complaints and has issued refunds when pushed
- The frustration it targets is real: juggling five AI tools and being the glue is a genuine pain
What doesn't
- The 'satisfied switcher' is the product's own CEO — an undisclosed ad dressed as a testimonial
- Promoter under active SEC fraud charges (Sept 2025, ~$112M alleged scheme with Tai Lopez)
- $8 buys starter credits; auto top-up defaults ON — documented user outcomes: $534 in 20 minutes, $600+ surprise charges
- Same low-ticket-hook playbook as Mentorbox, his earlier venture with years of BBB billing complaints
- 'No AI tool beats Supercool' — it's a wrapper around unnamed third-party models; the claim is unfalsifiable
- ~239 active Meta ads, ~26 new creatives a week — you didn't stumble on a post, you were targeted by a funnel
The catch
The $8 is a doorway, not a price. The product runs on credits, top-up is pre-checked, and real usage of what $211/month of dedicated tools delivers would cost multiples of $211 in credits. The comparison the post makes is the one thing the pricing model guarantees you can't have.
How to actually do it
- Before believing any 'I switched' post, search '<person> + <product> + founder'. Took us one search to find the CEO title
- Never leave auto top-up on with any credit-based AI tool — that's where the $534-in-20-minutes stories come from
- If you actually want to consolidate: pick the two tools you use weekly, cancel the rest. That's the honest version of cutting $211/month
- A tool that won't name the models it runs on is telling you the markup IS the product
If a testimonial moves you, the first check is always the same: does the person talking own the thing they're praising?
- Effort
- The 'one workflow' is real for toy projects; matching your current tools' output takes the same work plus a credit meter running
- Cost
- $8/month on the sticker; credits + default-on top-ups in practice — users document hundreds per month at real usage
- Stack
- Undisclosed third-party models behind a single interface — the opacity is the business model
- Confidence
- High
- Posted by
- a verified 'Ph.D. entrepreneur' page — who turns out to be the CEO of the tool he's 'switching' to
PARTLY TRUE "A $3 BILLION AI Berkshire Hathaway — want in?" — from a marketer under a federal deception order
The claimI'm launching the biggest company I've ever done after selling $100M+ online. Already raised $4 Million — want in? We're building a projected $3 BILLION AI 'Berkshire Hathaway' that buys 100 agentic AI companies (10 in the next 9 months). Also: build your own AI clone — free 3-day challenge at MakeMyClone.ai.
This one's real — that's what makes it dangerous. Anik Singal is a real person, Ugentic is a real SEC-registered Delaware company, the Tysons Corner summit is real, and the clone challenge is real. Now the gaps. In 2023 the FTC hit Singal and his company Lurn with a $14M judgment (partially suspended, $2.5M paid in refunds) and PERMANENTLY BARRED him from unsubstantiated earnings claims — and 'projected $3 BILLION... want in?' is exactly that genre of sentence. The '$4M raised' exists only in his own press release; the SEC-verifiable crowdfunding total is about $465K. The company is 18 months old, lost $2.9M on $2.2M revenue, and its two 'acquisitions' toward the 100-company empire are a newsletter and Singal's own previous company. To his credit, the actual SEC filings disclose the FTC order honestly — but the Facebook copy isn't the filing, and that gap is the whole game. The funnel underneath: free 3-day challenge → $37 VIP upgrade → high-ticket CloneIQ course → an equity 'want in?' on top.
What holds up
- Real, identifiable operator; real SEC-registered company with genuine Form C filings
- The crowdfunding raise, summit, and 3-day clone challenge all verifiably exist
- The Reg CF disclosure documents are honest — they even disclose his FTC order in the officer bios
- The underlying idea (AI clone of yourself for content) is real, buildable technology
What doesn't
- 2023 FTC order: $14M judgment, $2.5M refunded to ~1,900 customers, permanent ban on unsubstantiated earnings claims
- '$4M raised' is a self-issued press release; SEC filings verify ~$465K of crowd money
- '$3 BILLION projected' appears nowhere in any SEC filing — pure marketing number
- The 100-company roll-up currently consists of a newsletter and his own former company
- $0.10 illiquid Reg CF shares from a company running a $2.9M annual loss
- Long complaint trail on his programs: refused refunds, aggressive upsells — the FTC-ordered refunds confirm it wasn't anecdotal
The catch
Every layer of this feeds the next: free challenge → $37 VIP → high-ticket clone course → 'invest in me' equity ask → summit ticket. The product being sold at every step is proximity to a $3B number that exists only in the ad copy.
How to actually do it
- Want the AI clone? Build it: ElevenLabs voice clone (~$22/mo) + HeyGen avatar (~$29/mo) + a Claude project loaded with your own transcripts as the brain. ~$50-80/month total, one weekend of setup — no challenge, no upsell ladder
- Before ANY crowdfunding 'want in': read the actual Form C on SEC EDGAR, not the press release. Five minutes, free, and in this case it contradicts the pitch on the numbers
- When a promoter has a federal order about earnings claims, weight every number in the ad at zero until the filing confirms it
- Rule of thumb: real raised money has an SEC accession number; 'projected' money has a countdown timer
The clone tech itself is real and cheap to build yourself — that's the honest alternative, and we priced it out.
- Effort
- DIY clone: a weekend to stand up, a few evenings to tune the voice and load the brain
- Cost
- DIY: ~$50-80/month all-in. The funnel: free → $37 → high-ticket course → $1,000 minimum equity ticket at a 7% portal fee
- Stack
- ElevenLabs (voice) + HeyGen (avatar video) + Claude with your own content corpus (brain) — every piece off-the-shelf
- Confidence
- High
- Posted by
- a veteran internet-marketing guru's freshly minted secondary page (10K followers, a literal 'Test Post' in the feed)
PARTLY TRUE "I won the Skool Games 2x — my exact systems, FREE (for now)" — when 'winner' means top ten
The claimOnline coaches: I'm celebrating winning the Skool Games for the second time — and I'm giving away the exact strategies I use to get clients online. Free. For now.
The Skool Games are real — the monthly/quarterly competition on skool.com ranked by new recurring revenue, with Hormozi as the frontman and a Vegas trip for winners. The operator behind this reel is real too: an identifiable coach with a 4,300-member free Skool community and a paid program, whose bio matches the reel word-for-word. Here's the asterisk: Skool crowns the entire top TEN of every round as 'winners' — so 'I won the Skool Games 2x' most likely translates to 'I placed top-10 twice,' a real but much softer achievement than the phrase implies. And we couldn't independently verify even that: no public winners list or announcement names him; the claim lives only in his own bio. The private-event banner in the footage is consistent with a top-10 trip, so it's plausible — just unproven. The 'free (for now)' giveaway is a lead-magnet community whose job is booking sales calls for a paid program with no published price.
What holds up
- The Skool Games are a real competition with real, verifiable mechanics
- Real operator with a substantial, genuine Skool footprint — not an anonymous ad account
- The free community and its templates genuinely exist and cost nothing to join
- Top-10 placement is genuinely plausible for a community his size — the credential isn't absurd, just unverified
What doesn't
- 'Winner' is doing heavy lifting: Skool calls all ten top finishers per round 'winners'
- The 2x-winner credential is self-asserted only — absent from every public winners list we could find
- Brand-new FB page, single-digit engagement; the checkmark is paid Meta Verified, not notability
- 'FREE. For now.' is manufactured urgency on a lead magnet that was always going to be free — its job is booking calls
- 'Guaranteed' scaling language in the bio; no published pricing anywhere in the funnel
The catch
The whole pitch rests on one word — 'won' — that Skool's own format waters down to 'top ten.' Free bait, call-gated price, and a credential you can't check from the outside: the funnel is standard; the trophy is the marketing.
How to actually do it
- Before buying any Skool Games credential: ask WHICH round and WHAT placement, then ask for the announcement link. Top-10 finishes get announced — a real winner can produce the receipt in one message
- Take the free community if the templates are useful — just know you're entering a book-a-call funnel, and the 'for now' deadline is theater
- Never book the sales call before you know the price range. 'We'll discuss on the call' means the price flexes to your pain
- The actual playbook these funnels run — free value, community, call-gate — is itself the client-getting system being sold. You're watching the lesson for free
Not a scam — a real operator with an inflatable credential. Grade the systems, not the trophy.
- Effort
- Joining the free community: minutes. The 'exact systems' working for you: the same months of grind they took him
- Cost
- Free tier is genuinely free; the paid program's price is call-gated — assume high-ticket until told otherwise
- Stack
- A Skool free community + a book-a-call gate + a paid backend — the standard 2026 coaching-for-coaches ladder
- Confidence
- Medium
- Posted by
- a brand-new coaching-for-coaches page run by a real, identifiable Skool community operator
PARTLY TRUE "Google just made free what agencies charged $5,000 to build" — your own AI assistant in 5 minutes
The claimGoogle made custom AI assistants completely free — anyone can build their own without coding in under 5 minutes, the same thing agencies used to charge $5,000 to build. Read the comments for how.
The feature is real and genuinely good: Gemini Gems let you save a named assistant with custom instructions and uploaded knowledge files, no code, free on every Gemini tier — and five minutes is an honest estimate for a basic one. The rest is packaging. "Just made free" is stale: Gems have been free for well over a year. And the $5,000 anchor compares two different products — agencies charge that for a customer-facing chatbot embedded on your website with lead capture and integrations; a Gem lives inside your own Gemini account and talks to you, not your customers. The "read comments" close is the engagement-funnel tell: the post exists to farm comments, not to teach.
What holds up
- Gemini Gems are free on every Gemini plan, including the free tier
- Genuinely no-code: name, instructions, optional knowledge files, save
- Under 5 minutes is realistic for a basic Gem
- Real business value for repeat tasks — a saved specialist beats re-pasting prompts
What doesn't
- "Just made free" — Gems went free to all users in early 2025; the news is over a year old
- $5,000 agency comparison is apples-to-oranges: Gems aren't customer-facing website chatbots
- No lead capture, no site embed, no CRM hooks — the things the $5k actually buys
- "Read comments" = comment-farming funnel, the payoff is engagement for the poster
The catch
The $5,000 anchor. What agencies charge that for is a chatbot on YOUR WEBSITE talking to YOUR CUSTOMERS — embedded, branded, wired to lead capture. A Gem is a private assistant inside your own Gemini tab. It's free and worth building. It's just not the product in the price tag.
How to actually do it
- Go to gemini.google.com signed into any free Google account
- Click your profile avatar → Gems (Gem manager) → New Gem
- Name it, then write Instructions like a job description: who it is, what it does, output format. Use the rewrite wand if stuck
- Add Knowledge files (pricing sheet, FAQ, style guide) so it answers from your material
- Pick a Default tool if relevant, Save — it now sits in your sidebar as a reusable specialist
- Need the actual $5k product — a bot on your website answering customers and capturing leads? That's a different build: a chat widget wired to an AI backend with your knowledge base. Real project, not five minutes
Exact 5-minute Gem build steps below, plus what the $5k agency build actually adds so you know which one you need.
- Effort
- 5–10 min for a working Gem; a customer-facing website chatbot is still a real build
- Cost
- $0
- Stack
- Free Google account + Gemini Gems
- Confidence
- High
- Posted by
- an Instagram AI-content creator, cross-posted to Facebook
PARTLY TRUE "Build one AI brain that runs your day" — the $99 webinar with a fake-full seat counter
The claimStop collecting AI tools that do nothing — build one AI brain that runs your day: emails drafted, diary sorted, research done before your coffee's gone cold, even if you've never written a line of code. Live webinar, reserve your spot before they're full.
The concept is real — and we'd know, because building these is literally our business. The seller is a real, identifiable marketer (not anonymous), the stack he teaches is a genuine Claude-based setup with named AI 'employees' and job-specific skills, and a non-coder really can end up with a brain that drafts emails and prepares research. Now the other half: the ad's 'reserve your spot before they're full' urgency is a seat-counter widget on a webinar that reruns every two weeks (recording sold to no-shows); we could not find a single testimonial, review, or independent student result anywhere; and 'runs your day before coffee's gone cold' describes a mature, wired-in system — not what a beginner walks away with after a two-hour webinar. The funnel: free ad → $99 webinar → $2,000 done-with-you install (90 minutes of Zoom, a couple of custom skills, 30 days of email support, then you're on your own).
What holds up
- Real, named operator with a consistent public footprint — not an anonymous ad account
- The underlying product genuinely exists and works: a Claude-based 'brain' with named assistants and task skills is a real, buildable thing for non-coders
- The webinar is genuinely scheduled and hosted live (on a fortnightly cycle)
- Refund condition on the install tier ('working system or refund') is stated on the page
What doesn't
- 'Before spots are full' is a manufactured-scarcity counter on an evergreen fortnightly webinar
- Zero independent evidence anywhere: no testimonials, no reviews, no student results
- 'Runs your day' autonomy takes real wiring — calendar, email, context loading — that a webinar template doesn't deliver
- The install tier buys 2.5 hours of calls and 30 days of email support — after that, the 'brain' has no one maintaining it
The catch
The DIY gap. A template brain with no business context loaded, no integrations, and nobody maintaining it is a very smart notepad. What actually makes an AI brain 'run your day' is the loading, the wiring, and the ongoing operator — exactly the parts the webinar can't hand you.
How to actually do it
- If you're DIY-curious: the stack is real — a Claude subscription plus a structured note vault gets you a genuine starting brain for ~$20/mo, no $99 webinar required
- Before paying anyone for an 'AI brain': ask what BUSINESS CONTEXT gets loaded, what it integrates with, and who maintains it after day 30 — the answers are the product
- Ignore seat counters on evergreen webinars. If it 'sells out' every two weeks forever, it doesn't sell out
- Ask any seller for one named customer running the system 90 days later. No name, no wire
Logged with respect — this is our own category, sold honestly-ish at the concept level. The claims that needed flags got them.
- Effort
- The webinar is 2 hours; a brain that actually runs your day is weeks of loading and tuning
- Cost
- $99 webinar → $2,000 install tier → plus ~$20/mo for the AI subscription it runs on
- Stack
- Claude (desktop/Code) + a note vault + task skills — the same category of build we do for clients, which is why we graded the claims and not the ambition
- Confidence
- High
- Posted by
- a real NZ marketer selling a webinar-to-install ladder for a Claude-based AI brain
HOLDS UP "Vibe code a $600k database" — the tech-bro reel that's in on the joke
The claimA San Francisco tech bro brags you can vibe code a $600k database — posted with the caption "activate 'ragebait software developers' mode."
Second satire entry in the log, and like the last one, it's honest about what it is: the caption discloses the ragebait before you can even get mad. The creator is a real person building real AI tooling who runs a "nontechnical tech bro" comedy persona on the side — monetized through sponsorships and brand deals, not a course funnel; nothing is sold in the reel. There's no actual $600k-database incident (the number is comedic inflation of a real genre of posts about vibe-coding databases), but the failure mode being mocked is thoroughly documented: a vibe-coded women's-safety app left its database effectively public and leaked roughly 72,000 images including thousands of government IDs plus over a million private messages; a founder who bragged his SaaS was "100% AI-written, zero hand-written code" was hacked within days — API keys sitting in the frontend, paywall bypassed. The joke lands because the reality is worse than the punchline.
What holds up
- Satire is disclosed in the caption itself — "activate 'ragebait software developers' mode" — no claim is actually being sold
- Creator verified as a real AI builder (~138K followers) whose persona is the product; revenue comes from sponsorships, not viewer purchases
- The underlying premise checks out: documented cases of vibe-coded backends leaking databases and getting hacked within days of launch
What doesn't
- No real "$600k database" story exists — the number is invented for comedy, so don't quote it as a case study
- Ragebait format works precisely because plenty of viewers WILL take it at face value — the disclosure lives in the caption most people never read
The catch
Almost none — the rare reel where the bait is labeled bait. The only trap is secondhand: people who clip or repeat the claim without the caption turn disclosed comedy into undisclosed hype.
How to actually do it
- If you're vibe coding anything that stores customer data: the database config is the part you don't skip — real apps have leaked tens of thousands of IDs exactly this way
- Never ship AI-generated code with API keys in the frontend — it's the first thing attackers check, and it's how the "100% AI-built" SaaS got owned in days
- Enjoy the satire, but check captions before resharing 'claims' — disclosed comedy stripped of its caption becomes someone else's hype
Nothing to build — logged for the kernel, which is a genuinely useful warning wearing a costume.
- Effort
- 60 seconds to laugh; an afternoon to actually secure a vibe-coded backend
- Cost
- $0 — unless you skip the security pass, in which case: lawsuits
- Stack
- Any AI coding tool + a human who checks the locks before launch
- Confidence
- High
- Posted by
- a comedy tech-bro persona run by a real AI builder — the caption admits it's ragebait
HOLDS UP An open-source "agentic-first" CRM — the AI research agent IS the product, free and MIT-licensed
The claimA startup open-sourced the CRM it built for itself: a durable research agent that runs on its own schedule, keeps working after you close the browser, writes strong evidence straight to the record and hands weak evidence to a human — sandboxed, self-hosted, free.
The rare viral post that's basically all true — because it mostly just reads the project's own README, which undersells instead of overselling. The repo is real (MIT, launched July 31 by compliance startup Comp AI), and every architecture claim checks out against the code and docs: 18 authored tools plus 4 plain-English skill files, an explicit no-confidence-scores design, a sandbox with deny-all network egress and no database credentials, and a durable work queue so the agent schedules its own rechecks. The "20 stars" humility line was true at post time; it blew past 2,500 stars in three days. Even the warnings are honest: self-hosted, real setup (Bun, Docker, Google OAuth), promising-not-proven. One miss: the reel says "built on Claude" — the repo's default model is a different, configurable model; the reel creator added that detail, and built none of this himself.
What holds up
- Repo verified: trycompai/crm on GitHub — MIT license, created July 31, 2026, actively maintained, ~2,535 stars by Aug 3
- README matches the caption nearly word-for-word: 18 tools, 4 markdown skills, evidence-vs-suggestion split, sandboxed egress, durable queue
- Vercel's "eve" durable-agent framework is real (launched June 2026) and is exactly what the architecture leans on
- Comp AI is a real funded company (trycomp.ai, $2.6M pre-seed) that built this for its own sales team first
What doesn't
- "Built on Claude" is the reel's own embellishment — the default agent model in the code isn't Claude (it's swappable)
- The poster built nothing: it's a screen-record of the founder's launch post, riding a 24-hour star spike, with a comment-"CRM" DM gate for engagement
- Days old, ~zero production track record — the README says treat it as promising, not proven. Believe it.
The catch
Almost none on the product claim — the catch is the messenger. The reel account farms engagement off other people's launches, and the one detail it added ("built on Claude") is the one detail that's wrong. Read the source, not the reseller.
How to actually do it
- Skip the comment-"CRM" DM gate — the repo is public: search GitHub for the Comp AI open-source CRM
- Read the README's design sections first; the no-confidence-scores and sandbox reasoning are worth stealing even if you never deploy it
- Only self-host if you're comfortable with Docker + OAuth setup — this is a weekend project, not a signup button
- Treat it like any 3-day-old dependency: pilot on non-critical data until it has a track record
Logged as a genuine Build-It candidate: the durable-research-agent pattern (agent keeps its own notes, humans confirm weak evidence) rhymes with how we already run our own memory systems.
- Effort
- An afternoon to read and evaluate; a weekend to actually self-host
- Cost
- $0 software (MIT) + your server + model API usage
- Stack
- Bun, Next.js, Postgres, Docker, Google OAuth, a durable-agent framework, model of your choice
- Confidence
- High
- Posted by
- an AI-content account screen-recording another founder's launch post for comment-bait
PARTLY TRUE "3.4x ROAS — and that's just our average across 115+ B2B companies. If you don't get an ROI, you don't pay."
The claimA 4-person agency claims a 3.4x average ROAS across 115+ B2B companies, builds your entire client-acquisition system in 14 days — ads, funnel, automation, everything, no retainer — and guarantees you don't pay without ROI.
The agency is real and small: a named founder, a named 4-person team, and testimonial names that verify as real people with real companies on LinkedIn. That's more than most ads in this genre can show. But every headline number exists only on their own website and the founder's LinkedIn — no Trustpilot, Clutch, BBB, or G2 footprint anywhere, and the named case studies total three, not 115. The ad's landing page is cloaked (returns a 403 unless you arrive through the ad), and the "no ROI, you don't pay" guarantee has no published fine print anywhere we could find. In this industry, that guarantee almost always lives in the sales call, where "ROI" gets defined as qualified calls booked — not revenue closed — with a minimum ad-spend clause underneath it.
What holds up
- Real registered agency with a public team page and an active founder LinkedIn
- Three testimonial names verify as real people at real companies
- "Done-for-you funnel + ads in 14 days" is a genuinely deliverable service — the timeline itself isn't the lie
What doesn't
- 3.4x average, 115+ companies, "$95M+ generated": self-reported everywhere, verified nowhere
- Zero third-party review footprint — not even bad reviews; the brand is too new to have a paper trail
- Landing page cloaked from everyone but ad-clickers; guarantee terms published nowhere
- Founder's own posts claim 6–10x on his funnels while the ad averages 3.4x — pick a number
The catch
The guarantee is the hook, and it's unfalsifiable until you're on the sales call. Ask one question before any call: "Define the ROI in the guarantee, in writing." If the answer is 'booked calls' instead of 'closed revenue,' you've found the fine print.
How to actually do it
- Before hiring ANY "ROI guaranteed" agency: get the guarantee's ROI definition, minimum spend, and exit terms in writing before the call
- Ask for 3 client references you pick from their list — not the 3 they volunteer
- Weigh the honest math: a good funnel build is real value; 115-client averages from a team of four with no reviews is ad copy
Nothing to build — this is a vetting-checklist entry. The service category is legitimate; the claims as advertised are unverifiable.
- Effort
- A 20-minute diligence pass kills or confirms this class of pitch
- Cost
- These engagements typically run four figures up front regardless of the 'no retainer' framing
- Stack
- Meta ads + a funnel builder + CRM automation — the same stack any competent operator uses
- Confidence
- High
- Posted by
- the agency founder himself, running cold-traffic ads from a recently created page
PARTLY TRUE "Find Claude MCP" — run an AI-influencer side hustle on Fanvue with Claude and AI video
The claimUse Claude via MCP plus AI video generation to build and run a monetized AI-influencer account on Fanvue — comment "Claude" for the free step-by-step guide.
Here's the surprise: the plumbing is real. Fanvue ships an official hosted Claude MCP server — messaging, post scheduling, earnings analytics, mass DMs — and the AI video model flashed on screen is a real, current release. You genuinely can wire Claude to a Fanvue account today; this is not a fake-tool mashup. The bullshit is the economics and the omissions. The platform takes roughly 20%, typical new creators earn between nothing and a few hundred dollars a month in year one, and the famous AI-influencer success stories are agency-built outliers with pro teams. Unmentioned: Fanvue requires prominent AI disclosure and identity verification, and the social platforms you'd promote on actively demonetize mass-produced AI accounts — the promotion channel itself is the fragile part.
What holds up
- Fanvue's Claude MCP integration is official and documented by Fanvue itself — the reel's core tool claim is true
- The AI video model shown is real and currently available
- AI creators are a real, growing revenue segment on the platform — this niche exists
What doesn't
- Typical earnings: $0–500/month in year one; the headline cases are top-1% agency operations
- Fanvue mandates AI disclosure + ID verification; violations mean removal — the 'passive faceless empire' framing ignores this
- Meta actioned hundreds of thousands of unoriginal-AI-content accounts — your growth channel can vanish
- The poster is unverifiable and the 'free guide' is a comment-to-DM contact harvest
The catch
The guide-seller monetizes YOU, not fans. When the tools are real but the median outcome is near zero, the reliable business is selling the dream to the next cohort — which is exactly what a comment-gated free guide is for.
How to actually do it
- If you're genuinely curious, read Fanvue's own MCP and AI-content docs — no guide purchase needed, it's public
- Budget like a business: model the ~20% platform cut, content costs, and months of $0 before judging the hustle
- Disclose AI content everywhere the platform requires it — the account you save is your own
- Skip the DM funnel. Anyone selling a map to a gold rush earns more from maps than gold
Steps given for the honest version; nothing here for us to build.
- Effort
- Real daily content + audience work for months — the opposite of passive
- Cost
- Video-generation credits add up fast; plus your promotion time
- Stack
- Fanvue + its official Claude MCP + an AI video tool — all real, all public
- Confidence
- High
- Posted by
- a side-hustle creator with no verifiable footprint, gating a "free guide" behind comments
BS "70 AI Specialists for $57 (one-time)" — a Claude plugin that replaces $2,000–$5,000/month freelancers
The claimOne Claude plugin installs 70 AI specialists — copywriters, marketers, analysts, strategists — in 60 seconds for $57 (80% off the 'normal' $540), replacing freelancers who cost thousands a month.
The product exists — and that's the most generous thing we can say. The $57 buys a ZIP of 70 pre-written Claude skill/prompt configurations plus videos: a prompt pack, not software, not a plugin in any meaningful sense, and the sales page itself disclaims any Anthropic affiliation. The discount is manufactured: the ad anchors at "normally $540," while the seller's own sales page crosses out $285 — the fake original price changes depending on which door you walk through, and the 'launch sale' never ends. The seller's other properties carry review-site complaints: overcharges versus the advertised price, logins never delivered, refund friction. And the punchline: open-source collections of 75–100+ Claude subagents do the same job on GitHub for free.
What holds up
- The download is real — buyers do receive a pack of 70 prompt/skill files
- Claude skills and subagents are a real, useful mechanism — the underlying tech works
- Well-built prompt packs can genuinely speed up marketing tasks
What doesn't
- "Normally $540" in the ad vs. $285 crossed out on his own page — the discount is fiction either way
- A prompt pack styled as a "plugin" with "specialists" — it's text files, and free equivalents (75–100+ subagent repos) sit on GitHub
- Review-site complaints on the seller's other products: overcharging, undelivered access, refund fights
- "Replaces $2–5k/month freelancers" — no prompt file negotiates, strategizes, or owns outcomes
The catch
You're paying $57 for what GitHub gives away, from a seller whose discount math changes by traffic source. The 'employees for the price of lunch' frame is the tell — labor isn't what a text file does.
How to actually do it
- Want a team of Claude specialists? Install a free open-source subagent collection from GitHub — several maintained packs ship 75–100+ roles
- Read what's inside any pack before loading it into your agent — free or paid, prompts are dependencies
- If a discount matters to you, screenshot the price from two different devices before buying anything from an info-marketer
Steps given: the free path to everything this ad actually delivers.
- Effort
- 15 minutes to install a free subagent pack yourself
- Cost
- $0 for the equivalent open-source packs
- Stack
- Claude Code + any maintained open-source subagent/skills collection
- Confidence
- High
- Posted by
- a serial info-marketer who previously sold ChatGPT prompt bibles
HOLDS UP "Sorry, Claude is not designing this" — a designer's counter-claim to the AI-builds-everything genre
The claimThat stunning animated site everyone credits to AI? Claude can build it — but it isn't designing it. You still need a designer to make it work.
Refreshing entry in the log: a claim pointed AGAINST the hype, from someone with skin in the game. A designer shows a high-end animated product site and says the quiet part: the model can generate the code, but the taste — hierarchy, motion timing, restraint, the decisions that make it feel expensive — came from a human. This matches what our own audits keep finding from the other direction: the tools genuinely build, and the differentiator that survives is direction and judgment. It's light on specifics and it's also self-promotion for design services, but nothing claimed is false.
What holds up
- Consistent with every real 'AI built this' case we've audited: generation is real, art direction is human
- The poster is a working designer showing work, not selling a course or gating a guide
- No fake numbers, no manufactured urgency, no funnel tricks — a unicorn in this feed
What doesn't
- It's still marketing — the takeaway conveniently equals 'hire a designer'
- No process shown: a before/after of the AI's raw output vs. the directed result would have proven the point instead of asserting it
The catch
None worth the name. The only thing to remember is that 'AI can't do X' claims age fast — what stays true is that someone has to know what good looks like.
How to actually do it
- Use AI to build; budget human taste for what customers actually see — that's where trust is won
- When you see a jaw-dropping 'AI made this' demo, ask what the prompt-giver already knew about design
- If you can't afford a designer, steal structure from the best sites in your industry and let AI implement THAT
Logged as honest counter-programming; it's the same lesson our Built-It entries keep teaching.
- Confidence
- High
- Posted by
- a working UX designer showing his own portfolio piece
PARTLY TRUE "Go to Getlayers and copy a prompt — get the working 3D site 1:1, skip 99% of AI garbage"
The claimA library of battle-tested prompts where you copy one and get the exact 3D framework or interaction shown, 1:1 — a strict filter that cuts out 99% of useless AI output, saving days of prompt tweaking.
The product is real: a polished library of three.js/WebGL templates and prompts with a free tier and paid annual plans, built by a design agency with legitimate industry awards. Two problems. First, the reel never mentions that the 'discovery' being shared is the poster's own product — the agency built the site it's telling you to go find. Second, their own documentation contradicts the ad's core promise: prompts are validated against one specific frontier model, weaker models 'lose the motion timing, depth and layering,' and the true 1:1 exact-match experience is what the top paid tier's source-file downloads provide — not the copied prompt. The '99% filter' number appears nowhere but the ad. No independent reviews of the library existed anywhere we looked.
What holds up
- The prompt/template library exists, is polished, and has a genuinely usable free tier
- The agency behind it has real, verifiable design-industry credentials
- Curated, pre-tested prompts DO cut iteration time versus prompting from scratch — the honest kernel is real
What doesn't
- Undisclosed ownership: the 'helpful tip' is the maker advertising itself in third person
- Their own docs undercut '1:1' — results depend on using the exact model the prompts were tuned for, and exact-match is a paid-tier source-file feature
- 'Cuts 99% of AI garbage' is an invented number with no methodology anywhere
- Zero third-party reviews found; every claim about the product traces back to the product
The catch
The comment-«Prompt» gate feeds you into a subscription funnel ($99–$497/yr tiers) run by the reel's poster, and the DM-«Project» line feeds the agency's services. The tip is an ad wearing a trench coat.
How to actually do it
- Browse the library directly on its site — no comment-gate needed; the free tier shows you what's real
- If you try a prompt, run it on the exact model their docs specify — that's the honest version of '1:1'
- Never trust a discovery reel without checking who owns the thing being discovered — page transparency takes 30 seconds
Steps given — the free tier is genuinely worth a look for our own site work; the ad's framing is what earned the flags.
- Effort
- Minutes to try the free tier; the paid tiers are an annual commitment
- Cost
- Free tier; ~$99/yr for all prompts; ~$139–497/yr for source files
- Stack
- three.js/WebGL templates + a frontier model + your own design judgment
- Confidence
- High
- Posted by
- the agency that OWNS the site it's recommending — undisclosed self-promo
PARTLY TRUE Software engineers are finding "mega startup ideas" here — a free daily email with an idea from a $100M founder
The claimThis is how the new generation of software engineers finds mega startup ideas: a free newsletter that sends you one startup idea every day, sourced from a $100 million founder.
The product is real and better than most things we audit: a free, daily, well-regarded startup-ideas newsletter with 125,000+ subscribers. The hype is in the sourcing claim. By the team's own about page, the first 500+ ideas were written in-house by the two people who run it; they only recently began pulling ideas from a network of founders and operators. 'Every day from a $100M founder' is ad copy stretched over 'handpicked by us, sometimes with founder input.' The reel itself is the company advertising its own funnel.
What holds up
- Newsletter exists, publishes daily, and is genuinely free
- 125k+ subscribers; documented fast growth in a public beehiiv case study
- Ideas are written editorial content, not scraped filler
- Signup really does take seconds — no credit card, no upsell wall at the door
What doesn't
- 'From a $100M founder' daily is marketing gloss — most ideas historically written by the newsletter team
- The reel is the product's own ad dressed as an observation about 'the new generation of engineers'
- Free tier is the funnel: monetized via sponsor slots and a premium tier
The catch
This is an ad the product bought for itself, framed as a trend report. The only inflated part is attribution — the ideas mostly come from the two people who write the newsletter, not a daily parade of $100M founders. That makes it a normal media business with punchy ad copy, not a scam. Rare verdict here: the thing being sold is worth what it costs, because it costs nothing.
How to actually do it
- Want the ideas? Subscribe at their site — email in, done. It is actually free.
- Want better startup ideas than a newsletter can give you? Mine your own customer complaints and niche communities — ideas you have unfair context to execute beat ideas 125k other people also just read.
- Want what they actually built? Copy the engine, not the ideas: daily short vertical videos with one bold on-screen claim, every one pointing at a single free email capture, sponsors once the list is real. Their beehiiv case study lays out the numbers.
Nothing to build — it's a subscription link, and it's safe. Logged the honest read: subscribe if you want idea flow, but the more valuable move for any small firm is copying their growth engine, which is publicly documented.
- Effort
- 10 seconds to subscribe; copying their growth engine is a months-long content grind
- Cost
- $0 — ad-supported, optional premium tier
- Stack
- Any ESP (they use beehiiv) + relentless short-form video, if you copy the model
- Confidence
- High
- Posted by
- the newsletter's own Facebook page — this reel is a self-ad
HOLDS UP The "best Claude Code skills for animation" — GSAP, Three.js, motion design, creative coding
The claimFour GitHub skill repos — official GSAP skills, Three.js skills, an official LottieFiles motion-design skill, and a creative-coding plugin — give Claude Code better context for building animation and motion work.
The rare AI listicle that checks out. All four repos exist under the exact names shown, and the two biggest claims — "official" — are true: the GSAP skills repo (12.8k stars) is published by GreenSock itself, self-described as "Official AI skills for GSAP," and the motion-design skill is published by the LottieFiles org. The creative-coding plugin’s rebrand note on the slide matches its README word-for-word and shipped a new version the week we checked. The premise is vendor-endorsed, not influencer-invented: when the maker of an animation library publishes skills to teach AI agents its correct current API, "better context" is exactly what you get — these load real, current knowledge that models routinely get wrong from stale training data. Only bullet-level embellishments: the Three.js pack has no particle-effects content despite the slide, and the LottieFiles skill is motion *principles* (timing, easing, choreography), not Lottie export tooling.
What holds up
- greensock/gsap-skills: real, org-owned, "Official AI skills for GSAP," 12.8k+ stars, actively updated
- LottieFiles motion-design-skill: real and genuinely first-party — a taste-and-timing principles library that works with any animation system
- Three.js skill pack: 10 real skills covering scenes, cameras, GLTF, lighting
- Creative-coding plugin: active solo project, shipped a new version within days of the post; slide text matches the README verbatim
- All four use Anthropic’s standard skill format and install cleanly
What doesn't
- "Particle Effects" on the Three.js slide: zero particle content in any of its 10 skills
- The LottieFiles skill won’t touch Lottie/dotLottie files — it teaches motion principles, not export tooling
- The Three.js pack’s own README clone command points at the wrong repository — install by copying its skills folder instead
The catch
Almost none — which is the story. The only trap is expecting two of the slide bullets (particles, Lottie export) that the repos never promised.
How to actually do it
- Install the official GSAP skills first — biggest payoff, first-party maintained
- Add the LottieFiles motion-design skill for taste: timing, easing, and choreography guidance on any stack
- Grab the Three.js pack only if you build 3D — and copy the skills folder manually (its own install command is broken)
- Vet third-party skills like any dependency: read what’s in the folder before you load it into your agent
This one goes in the toolbox. Verified worth adopting for real client site work.
- Confidence
- High
- Posted by
- a working Flutter developer sharing free dev tooling
PARTLY TRUE "5 free repos that run a faceless video channel — no camera, no editing, posts while you sleep"
The claimFive free GitHub repos — an agentic video studio, a fast video-diffusion model, a talking-avatar generator, a caption burner, and a clip cutter — chain together into a faceless video channel that posts while you sleep.
Rare for this genre: every repo is real and every number on the slides checks out (we verified all five live on GitHub — the flagship agentic video studio really is one of the fastest-growing repos of 2026 at 44k+ stars, and it really can produce a finished faceless short for cents-to-dollars in API costs). The ending is the fiction. Not one of the five uploads, schedules, or posts anything — the flagship’s own README says it renders local video files and "requires active agent interaction." The clip-cutter was abandoned after a single day of commits and misfits the stack anyway (it clips long videos a generated-b-roll channel wouldn’t have). The diffusion model’s "100× faster on your own GPU" is a real benchmark — measured on RTX 5090/H100-class hardware with custom CUDA kernels. And if you did wire up the missing uploader, YouTube’s "inauthentic content" rules and TikTok’s "unoriginal content" flag are aimed directly at unattended mass-produced AI output.
What holds up
- All five repos exist under the exact names shown; star counts were accurate within a week’s drift
- The agentic video studio genuinely builds finished faceless shorts via an AI coding agent for $0.15–$3/video in API costs
- The diffusion speedup and talking-avatar tools are legitimate academic-grade projects
- The caption tool does exactly what the slide says
What doesn't
- Zero of the five repos post, schedule, or upload — "posts while you sleep" is covered by nothing in the stack
- Realistic burden: a ~$1,600+ GPU or cloud rental, CUDA dependency builds, 14B-model downloads, and babysitting each render
- The clip-cutter’s entire commit history is one 34-minute day — abandoned, and the wrong tool for this pipeline
- The talking-avatar tool degrades after ~1 minute of video by its own docs
- YouTube demonetizes "inauthentic" mass-produced content; TikTok’s unoriginal-content flag blocks Creator Rewards — the business model dies at the payout step
- The slideshow is top-of-funnel for paid 1:1 calls
The catch
The slides sell you the assembly line and skip the fact that there’s no shipping department — and that the stores it would ship to have policies against the product.
How to actually do it
- Treat the stack as a production accelerator: human script and editorial choices, AI does the rendering grunt work
- Start with the caption burner and clip tools on content you actually filmed — lowest cost, immediate payoff
- If you try the generative tools, budget for GPU reality: cloud rental beats buying hardware for a test
- Post manually with AI-content disclosure enabled — monetization policies reward originality, so add your voice, your face, or your take
Used honestly, these tools are a genuinely fast assembly line you operate — not an autopilot.
- Confidence
- High
- Posted by
- a prompt-selling TikTok creator who funnels to paid 1:1 calls and app builds
BS "6,000 pieces of content a month from 2–4 hours" — blasted through "hundreds of dedicated fan accounts"
The claimRecord 2–4 hours a month and their system turns it into up to 6,000 posts distributed through "hundreds of dedicated fan accounts" — the same strategy that built Alex Hormozi and Iman Gadzhi, already used by Daniel Priestley, Russell Brunson, Brandon Dawson, and MambaFX.
The mechanic they sell is the problem. Their own review-site description admits distribution runs on "hundreds of fan/sub accounts" posting the same content — which is precisely the coordinated inauthentic behavior TikTok, Meta, and YouTube purge on sight (TikTok removed 9,700+ accounts in one network sweep; Meta pulled ~500,000 spam profiles in its 2025 crackdown). The company itself is a ghost: the domain is about 18 months old and hidden behind a privacy proxy while its LinkedIn claims "founded 2019," no founder or legal entity is named anywhere, and there is zero independent evidence any of the name-dropped gurus hired them — the names appear only on their own page. The Hormozi comparison is false on mechanics: he built an in-house media team publishing on his OWN channels, not throwaway account networks.
What holds up
- Content repurposing/clipping is a real, legitimate tactic — one long recording genuinely becomes dozens of clips
- The company exists and some of its view-count screenshots may be real numbers — spam networks do generate views
- Long-form → clips → omnipresence is a sound strategy when run on accounts you own
What doesn't
- Their own public materials describe a mass fan-account network — textbook platform-TOS violation on every network they list
- Anonymous operation: no founder, no address, no legal entity, no terms or privacy policy on the funnel page
- Domain registered Jan 2025 behind a privacy shield; LinkedIn claims "founded 2019" — the timeline doesn’t reconcile
- Zero independent evidence for the Priestley / Brunson / Dawson / MambaFX client claims
- "THEY SHOCKED @MRBEAST" hook: no MrBeast connection exists anywhere in their materials
- Views are not leads — spam-network impressions don’t become customers who trust you
The catch
You’d be paying an anonymous company to run a coordinated fake-account network in your name. When the platforms’ regular purges hit it — and they do, at scale — the views evaporate and the bans and spam association land on YOUR brand, not theirs.
How to actually do it
- Record the long-form content exactly as pitched — that part of the model is right
- Have a small team or an AI-assisted pipeline you own cut it into clips
- Publish on accounts YOU control, where the audience and algorithm equity compound to your brand
- Judge any content vendor by leads and revenue, never raw views — and walk away from anyone whose distribution method can’t survive a platform policy read
The honest version of this offer exists and works — it just doesn’t involve fake accounts.
- Confidence
- High
- Posted by
- Growthopia, an anonymous "omnipresence" content agency advertising to coaches and experts
BS "$0 to $400M in 3.5 years" — the done-for-you AI business
The claimHis AI business went $0 to a $400M valuation in 3.5 years, runs itself on 4 hours a day from Greece, and his team will build YOU one of these "AI Arbitrage" businesses — no college, no technical skills, just monthly recurring revenue while AI does the heavy lifting.
The person is real — Jordan Lee co-founded AI Acquisition (UK) — and that's roughly where real ends. The $400M valuation is self-declared: the company has zero outside funding, which means no priced round and no independent valuation event, and the number drifts between his own placements ($300M in one ad, "multi-9-figure" in another, $400M here). The only press is pay-to-play. What the funnel actually sells: a $97 "blueprint" that ladders to $5K, $10K, and $25K coaching tiers — where you still do all the prospecting, selling, and delivery yourself. An independent audit of Trustpilot sentiment ran ~92% negative, including buyers reporting six figures spent and nothing earned. And the whole "done-for-you business that runs itself" genre is a repeat FTC target: DK Automation, Automators AI (owners banned from the industry), Ascend Ecom, Click Profit — same script, same finding: most buyers lose money.
What holds up
- Jordan Lee is a real person and AI Acquisition is a real company — the persona checks out
- The funnel structure ($97 front end → $5K–$25K coaching tiers) is documented by independent reviewers
- The valuation appears ONLY in his own ads and pay-to-play press releases — no funding database, no priced round, no independent source
- Independent Trustpilot audit: ~92% negative sentiment, incl. "$100K spent, not one cent made"
- FTC has taken repeated federal action against the DFY/automation genre, with bans and eight-figure judgments
What doesn't
- Self-declared valuation as the headline hook — no outside investors means nobody independent ever priced it
- Lifestyle-porn framing: Greece, 4 hours a day, "business runs itself"
- Comments effectively silent on a 23K-view ad — engagement managed
- Classic VSL ladder: free presentation → sales call → five-figure close
- Income claims sourced from a cherry-picked survey of his own customers
The catch
The ad sells a $400M success story and a business that runs itself. The valuation exists only in his own ad copy, the "business they build you" is a $5K–$25K coaching package where you do the actual work, and the buyer receipts run 92% negative. The cruel irony: the underlying idea — helping everyday businesses implement AI tools that already exist — is a genuinely real business. It's just a job, not a money printer, and nobody sells you a job for $25K.
How to actually do it
- The honest version: local businesses really do pay for AI implementation — as a service you sell and deliver like any trade. Real work, real revenue, no passivity anywhere
- Before paying anyone to "build you a business": search their name + Trustpilot and read the one-star pages first; then search the genre + FTC
- Valuation test: no outside investors = no independent valuation. Ask who priced it. (There's no answer)
- "Runs itself" test: ask exactly who answers the customers, does the delivery, and handles refunds in month three. If the answer is "AI," ask to watch it do that
- Want the real thing? Pick one deliverable skill — intake bots, memory systems, follow-up automation — deliver it to one local business for a fair fee, and compound. That's the whole secret, and it's free
The kernel is real and worth doing honestly — here's the difference between the fantasy and the actual business, plus the checks to run before anyone pays a guru five figures.
- Confidence
- High
- Posted by
- Jordan Lee, co-founder of AI Acquisition, running paid video ads
PARTLY TRUE Violoop: 'Two cables. No coding. Running in five minutes' — a $300 box that turns any PC into a 24/7 AI employee
The claimA palm-sized plug-and-play device (HDMI + USB, ~$300) watches your screen, physically controls keyboard and mouse, and turns any ordinary PC into an always-on AI 'digital employee' that automates any app — no coding, running in five minutes.
The hardware is real and the architecture is genuinely clever: HDMI screen capture plus real USB keyboard/mouse simulation means it can drive any application — even ones with no API — and a physical confirmation key for risky actions is a better safety design than most software agents have. The five-minute setup claim is plausible. What's oversold is everything after minute five: an 8B on-device model handles only routine tasks (complex work routes to your own cloud API key, thinning the privacy pitch), screen-driven agents in 2026 still fumble multi-step workflows regularly, and 'zero detection, zero bans' is bot-detection evasion marketed as a feature. Add that it's crowdfunded hardware from a startup months old, and '24/7 digital employee' is a demo reel, not a deliverable.
What holds up
- Real device, real funding (seed+angel, ~$M USD), Kickstarter launched April 2026 at ~$300
- Hardware HID + HDMI approach genuinely works with API-less apps
- Physical confirmation key for destructive actions is a legitimately good safety design
- No software install on the host PC; doesn't consume its CPU/GPU
- BYO API key (Claude/OpenAI/Gemini) confirmed on the official site
What doesn't
- '24/7 digital employee' oversells what screen-vision agents reliably do in 2026
- On-device 8B model = routine tasks only; hard tasks hit your cloud API key and bill
- 'Zero detection, zero bans' openly advertises ToS-evasion of platforms that prohibit automation
- Crowdfunded first-gen hardware from a months-old company — delivery and support risk
- No meaningful base of independent long-term user reviews yet
The catch
The pitch quietly swaps 'set up in five minutes' for 'working reliably in five minutes.' Setup is fast; dependable multi-step automation is the unsolved part — and their own press coverage cites an AI agent deleting 2,000 emails as the exact failure mode the confirm-button exists for. You're buying a promising v1 gadget, not an employee.
How to actually do it
- Want the outcome now: put Claude Code (or another frontier agent with computer use) on a spare mini PC — same class of automation with a far stronger brain, for the cost of hardware you may already own
- Scope it to tasks with low blast radius: data entry, report pulls, file organization — never irreversible actions unattended
- If Violoop itself appeals, wait for retail availability plus 90 days of independent user reviews before ordering; treat crowdfunded hardware as a preorder gamble, not a purchase
- Whatever runs the automation: check the terms of service of each app you automate — hardware-level 'undetectable' input doesn't change what a platform's rules allow
- Budget the true cost: device price + cloud API usage for any non-trivial task
The honest path to the same outcome exists today without waiting on crowdfunded hardware — laid out below, plus the sane way to evaluate Violoop if the concept appeals.
- Effort
- Setup: minutes, as claimed. Reliable unattended workflows: weeks of tuning, on any platform
- Cost
- ~$300 device + your API-key usage; DIY route: $0 software on hardware you own
- Stack
- Violoop (or: a mini PC + Claude Code / a computer-use agent)
- Confidence
- Medium
- Posted by
- a venture-funded Chinese AI-hardware startup advertising its own device
PARTLY TRUE The 'AI Inbound Client System' that turns feast-or-famine agencies into $5–15K retainer machines
The claimYour agency doesn't have a scaling problem, it has a system problem — and their done-with-you 'Inbound Retainer System' installs the offer, paid-ads acquisition, sales frameworks, and delivery SOPs that produce $5K–15K/month recurring retainers.
The business advice inside the ad is sound and free: retainers beat projects, and predictable inbound beats referral roulette. The company is real, with hundreds of positive reviews and students who do report landing retainers. What earns the discount: the 'AI' in the video title is decoration on a standard paid-ads-plus-funnel-plus-sales-scripts model, the price is never stated (this is high-ticket coaching), and the glowing review wall skews heavily toward people in their first days of the program — enthusiasm about onboarding, not evidence of outcomes. The system is real; whether YOU need to buy it is the actual question the ad is built to keep you from asking.
What holds up
- Retainer-ization of agency services is a proven, legitimate model
- Real company with 700+ reviews; some students report actual retainer wins
- Done-with-you coaching genuinely beats passive course libraries for implementation
- The diagnosis — feast/famine comes from referral dependence — is accurate
What doesn't
- 'AI' appears in the title, not the method — the system described is ads + funnels + scripts
- Price undisclosed anywhere in the funnel; that's a feature of the funnel, not an oversight
- Review wall is heavy on day-one and week-one students, thin on 90-day outcomes
- 'We run this in our own company' is circular proof — their evidence the ads work is that their ad reached you
- Results depend on ad spend the pitch doesn't budget for
The catch
The entire ad is the product. A pitch teaching you to sell high-priced retainers is itself a high-priced retainer sale — you are watching the system be run on you. That's not disqualifying, but price it in: the model they teach is public knowledge, and what you'd actually be buying is implementation speed and accountability at an undisclosed high-ticket price.
How to actually do it
- Repackage your services into 2–3 monthly retainer tiers anchored to a business outcome, not hours or deliverables
- Take it to your existing clients first — the fastest retainer is converting a project client to a monthly engagement at the point of delivery, when trust peaks
- Build one inbound channel before touching paid ads: consistent authority content plus a single clear offer page
- If you evaluate any coaching program: get the total price and refund terms in writing on the first call, and ask for 3 references at YOUR revenue level who are 6+ months out
- Filter their reviews to ones describing results after 90+ days; ignore onboarding-week praise entirely
- Decision rule: pay for coaching only to buy speed and accountability — never information. The information is free.
Extracted the working core — the retainer conversion play — as steps anyone can run without buying anything, plus a vetting protocol if you do want the done-with-you version.
- Effort
- Retainer repackaging: a focused week. Inbound flywheel: 90+ days either way
- Cost
- DIY: ~$0 beyond ad spend. Coaching route: undisclosed high-ticket + ad budget
- Stack
- An offer doc, a booking page, one content channel; ads only after organic proof
- Confidence
- Medium
- Posted by
- a video-agency coaching company advertising to 6–7 figure agency owners
PARTLY TRUE "Say goodbye to manual SEO" — a full agency campaign in 15 minutes
The claimAn agency owner built a real SEO campaign in 15 minutes — keyword research, content, backlinks, and reports — with instant keyword clustering, guaranteed DR45+ PR backlinks, auto-publishing content, and white-label reporting.
The product behind the ad (Adaptify) is real and works — 500+ agencies, verified reviews across Trustpilot, G2, and Capterra confirm it genuinely automates keyword clustering, content generation, auto-publishing, PR-outreach backlinks, and white-label reports. What's oversold: 15 minutes buys you a campaign *plan* in a dashboard, not a campaign. Content still needs human review (users report accuracy issues), backlinks arrive over weeks via journalist pitching with variable placement rates, and rankings take months. And the 'try it free' ad hides retainer-level pricing that isn't published anywhere.
What holds up
- Real company: founded 2023, TinySeed-backed, 500+ agency customers
- Verified users confirm the time savings — 'a day's work in 30 minutes'
- Backlinks come from real journalist/guest-post outreach, placements reviewable
- Users independently confirm DR40+ link placements
- White-label reporting and auto-publish are real shipped features
What doesn't
- '15 minutes' = the setup and plan, not a live campaign — the Demo Cut
- 'Guaranteed' backlinks depend on outreach placement rates that vary by niche
- Users report content accuracy issues and slow external backlink results
- Pricing not public — reviewers peg it around $499–999+/mo behind the 7-day trial
- Scaled guest-post link building lives in Google's gray zone; the ad never mentions it
The catch
The ad sells a 15-minute campaign and a 7-day free trial. The reality is a $500–1,000/month retainer-class tool where the 15 minutes gets you a plan, humans still review the content, and the 'guaranteed' links show up on outreach timelines — weeks, not minutes. Good tool; dishonest clock.
How to actually do it
- Take the 7-day trial, but connect one real client site — not a test domain
- Generate the campaign plan on day one; that is the genuine 15-minute part
- Publish 2–3 articles to a low-stakes blog and grade them against your current writer's output
- Before trusting 'guaranteed,' ask sales for last month's actual placement rate and 5 sample placements — verify the DR yourself in Ahrefs or a free DR checker
- Get the real monthly price in writing and compare it to your current content + link vendor spend
- Decision rule: if it holds your quality bar and saves 10+ hours/month at that price, keep it; if not, a Claude-based content pipeline plus a vetted link vendor covers the same ground for less
The working core is worth a structured trial for any agency doing SEO delivery. Wrote the exact evaluation path below — decision rule included, so the trial ends in a keep-or-kill call, not a vibe.
- Effort
- 15 min to a plan; weeks for links; 3–6 months for ranking results
- Cost
- ~$499–999+/mo per reviewer reports; price not published
- Stack
- Adaptify (or: Claude content pipeline + link vendor + rank tracker)
- Confidence
- High
- Posted by
- a venture-backed SEO SaaS running paid ads to agency owners
HOLDS UP Turn your notes app into a 'second brain' with Claude + Obsidian
The claimConnect Claude to an Obsidian vault and it becomes a 'second brain' — an AI that reads, searches, links, and organizes your entire body of notes.
This one holds up. Obsidian stores notes as plain markdown files on your own machine, and Claude Code (or Claude Desktop via MCP) can read, search, write, and cross-link the whole vault directly. Multiple independent write-ups between February and June 2026 reproduce the same result with the same 20–30 minute setup, and the entire stack costs nothing beyond a standard Claude subscription. A rare viral claim where the capability is exactly as advertised — the only thing the video can't give you is the note-taking habit that makes a second brain worth having.
What holds up
- Obsidian vault = plain local markdown; Claude reads and writes it directly
- Claude Code needs zero plugins — point it at the vault folder
- Independently reproduced by multiple writers, Feb–Jun 2026
- Notes stay local; nothing uploaded beyond normal Claude usage
- Setup consistently reported at 20–30 minutes
What doesn't
- Silent video — the exact on-screen method can't be verified from outside the platform
- A second brain is only as smart as the capture habit feeding it — the demo skips that part
The catch
The creator funnels viewers toward a ~$94/month community and done-for-you builds — to teach a 30-minute setup that is documented for free all over the internet. The capability is real; the paywall wrapped around it is the hype.
What we did
- Install Obsidian (free, obsidian.md) and create a new vault — one folder of markdown files
- Inside the vault create folders: Inbox, Daily, Projects, People, Resources, Maps, Templates
- Create CLAUDE.md at the vault root describing the folder rules and how you want notes filed, linked, and summarized — this becomes Claude's operating manual
- Install Claude Code (docs.claude.com), open a terminal, cd into the vault folder, run: claude
- First prompt: 'Read CLAUDE.md, index every note, and write Maps/overview.md linking the major themes'
- Prefer an app over a terminal? Claude Desktop → Settings → Extensions → Filesystem → add the vault folder — same result
- Daily use: dump raw thoughts into Inbox, then ask Claude to file, link, and summarize them per CLAUDE.md
Verified the full stack against current docs and independent reproductions, then built a ready-to-drop-in vault starter kit (folder structure, CLAUDE.md operating manual, capture templates). The steps below work standalone with no kit and no course.
- Effort
- 20–30 min setup; the ongoing capture habit is the real work
- Cost
- $0 beyond a Claude plan (~$20/mo); Obsidian is free for personal use
- Stack
- Obsidian + Claude Code (or Claude Desktop + filesystem MCP)
- Confidence
- High
- Posted by
- an ex-Meta engineer turned AI-automation creator, ~174k followers
PARTLY TRUE "This ChatGPT prompt makes SEO 100x easier" — let AI interview you, then rank
The claimPaste one prompt into ChatGPT and it interviews you with 10 questions (answered by voice) before writing your blog post — making SEO "100x easier" and you "99% more likely to rank."
The technique is genuinely good; the numbers are invented. Having AI interview you first, then write from your answers, is the right way to get firsthand experience into a post — and firsthand experience is exactly what Google's helpful-content systems reward. But "100x easier" and "99% more likely to rank" are decoration: no such statistics exist, E-E-A-T isn't a direct ranking factor, and one prompt does zero keyword research, zero links, zero technical SEO. Keep the interview, ignore the math.
What holds up
- Interview-first drafting works in any capable chat AI today, free tier included
- Google's own docs reward original info and firsthand experience
- Voice-mode answers are a real, low-friction way to capture expertise
- One-question-at-a-time pacing produces far better raw material than a single dump
What doesn't
- "99% more likely to rank" — fabricated statistic, sourced from nowhere
- "100x easier" — unfalsifiable hype math
- E-E-A-T guidelines inform quality raters; they are not a direct ranking dial
- Prompt skips keyword research, search intent, internal links, technical SEO entirely
- Reframed basic: this is standard journalism interviewing, rebranded as an AI hack
The catch
The prompt is real and useful — the SEO promise stapled to it is not. Ranking still depends on keyword targeting, site authority, and links. This prompt improves one input (content quality); it does not touch the other 80% of what decides whether you rank. The invented "99%" stat exists to make a decent writing technique go viral.
What we did
- Pick one real question a customer has actually asked you — that's the post topic
- Paste the upgraded prompt: "I'm writing a blog post answering: [question]. First ask me the target search phrase and who's searching it. Then interview me — max 10 questions, one at a time — digging for firsthand stories, numbers, mistakes, and opinions a competitor couldn't fake. I'll answer by voice; wait for each answer. Then draft the post: answer the question in the first 100 words, use my exact stories, add H2s phrased as related questions people search, and end with 3 internal-link suggestions and a meta description under 155 characters."
- Answer out loud with specifics — client names redacted, numbers kept; specifics are the whole value
- Edit the draft for voice, add one original image or screenshot (experience signal Google explicitly looks for)
- Before publishing, verify the target phrase actually gets searched — free check in Google Keyword Planner or by typing it into Google and reading the autocomplete
Extracted the working core and upgraded it: an interview-to-post workflow that also bakes in the SEO layer the original skips (target query, search intent, headers that match the question, internal-link prompts). Full upgraded prompt in the steps — works in Claude or ChatGPT.
- Effort
- 20–30 min per post, honestly stated for once
- Cost
- $0 — works on free tiers of Claude or ChatGPT
- Stack
- Claude or ChatGPT (voice input optional), Google Keyword Planner (free)
- Confidence
- High
- Posted by
- an SEO educator with a verified account
PARTLY TRUE “F*ck Notion” — connect Claude to Obsidian once and “let it run your life”
The claimDitch Notion and the 400 screenshots you'll never open: connect Claude to Obsidian one time and let it run your life — and steal the template the poster says they've used for three years.
The core is real and genuinely good: pointing Claude at an Obsidian vault of plain markdown is a proven, free, widely replicated second-brain pattern — one prompt and a folder, no purchase required. The wrapper is a paid ad. “Let it run your life” oversells note read/write into autonomy that doesn't exist without extra plumbing, and the “template” is a lead magnet for a funnel — in this genre, gated templates routinely arrive as an email sequence instead of a working vault.
What holds up
- Claude ↔ Obsidian works today — Desktop filesystem connector or Claude Code in the vault folder
- Vault is plain markdown you own; no lock-in, works offline
- One-time setup is honest — roughly 15 minutes
- The pattern is replicated by many independent creators, several with truly free templates
What doesn't
- “Run your life” = reads and writes notes; nothing runs unattended out of the box
- It's a paid ad (‘Learn more’ funnel) — the template exists to capture you, not to help you
- Claude–Obsidian integration is under two years old; a ‘3-year’ Claude-connected template can't exist as framed
- Documented buyer complaints in this genre: ‘free template’ gates that never deliver the files
The catch
You are the product being organized. The connection being sold costs $0 and one prompt — the ad exists because a template funnel needs your email, not because the setup is hard.
What we did
- Install Obsidian free from obsidian.md — ‘Open folder as vault’ on any folder
- Route A: Claude Desktop → Settings → Connectors → Filesystem → add the vault folder
- Route B: terminal in the vault folder → type ‘claude’ (Claude Code reads CLAUDE.md automatically)
- Add a CLAUDE.md telling it your folder map and standing behaviors — this is the whole ‘template’
- First prompt: ‘Read CLAUDE.md, interview me, and build out my projects, clients, and people notes’
Built the template they're selling and gave it away: a Claude-ready Obsidian starter vault (CLAUDE.md operating instructions, daily/client/project/person templates, inbox workflow) plus the exact connection steps.
- Effort
- 15–30 min, one time
- Cost
- $0 beyond the Claude plan you already have
- Stack
- Obsidian (free) + Claude Desktop or Claude Code + plain markdown
- Confidence
- High
- Posted by
- a verified productivity-brand ad account running a template funnel
PARTLY TRUE The "AI client acquisition machine" — calls booked while you sleep, you just show up and close
The claimWe install an AI system for agency owners: AI ads and content attract your ideal client, stages 1-4 run fully automatically with zero involvement, a nurture agent runs 14 touchpoints over email and SMS, and qualified calls land on your calendar while you sleep.
The pipeline is real and unremarkable — ad platform, CRM automation, AI-written follow-up sequences, and a booking calendar have been wired together for years; AI just writes the copy faster. What's oversold is "fully automatic, no involvement": the machine runs on paid ads, which means a monthly ad budget the reel never mentions, plus ongoing creative testing and offer positioning that no installer automates away. And the 14-touchpoint SMS nurture has a legal ceiling: US business texting requires A2P 10DLC registration and consent under TCPA — send automated texts to people who didn't opt in and the fine risk is yours, not the installer's. Real plumbing, silent fuel costs, unnamed compliance.
What holds up
- AI-drafted ads, nurture sequences, and auto-booking are standard, working tech
- Pre-qualifying before a human call genuinely raises close rates
- Keeping the human on the sales call is honest — full-AI closing is worse
- 2-3 hrs/week of oversight vs 20 hrs of manual prospecting is a fair before/after
What doesn't
- "Genuine ads" = paid ads; the required monthly ad spend is never stated
- "Stages 1-4 fully automatic, no involvement" ignores creative testing and offer work
- Automated SMS at scale requires A2P 10DLC registration and TCPA consent — never mentioned
- "Booked while you sleep" depends on the offer and ad budget, not the plumbing
The catch
The system is the cheap part. The reel sells the plumbing and hides the fuel: ad spend every month, deliverability upkeep, SMS compliance registration, and the offer positioning that actually decides whether anyone books. The installer gets paid either way.
How to actually do it
- CRM + automation platform with pipelines, email/SMS, and calendar booking built in (GoHighLevel-class tool, roughly $100-300/mo)
- Register your business for A2P 10DLC before any automated SMS — carrier registration plus documented opt-in consent; skipping this risks TCPA liability
- Draft the ad angles and 14-touch nurture sequence with Claude, in your voice, against your real offer
- Meta ads with a realistic test budget — expect four figures monthly before the calendar looks like the demo
- Human takes every sales call; review the sequence's reply rates weekly and rewrite the weakest touch
Broke down the honest stack, real running costs, and the compliance step the reel skips.
- Effort
- 1-2 weeks to stand up; ongoing weekly tuning forever
- Cost
- $150-400/mo software + ad spend on top; DFY installers charge a premium for the same parts
- Stack
- CRM/automation platform, Meta ads, Twilio-class SMS with A2P registration, Claude for copy, calendar booking
- Confidence
- High
- Posted by
- an agency-growth operator selling done-for-you AI installs
PARTLY TRUE "Anthropic spent $2.3M on a loop manual" — 7 secret principles that make agents earn
The claimAnthropic spent $2.3M and 4 years building a 7-principle "loop manual" — 1000+ hours of lectures — that turns any agent into a multi-task assistant that earns instead of falls.
The material is real and it is the best agent-building guidance published anywhere — but every number in the post is invented. There is no $2.3M manual, no 4-year development saga, no lecture vault. What exists is a handful of free engineering essays on Anthropic's public blog (Building Effective Agents, Effective Context Engineering, the Agent SDK guide) that the community later nicknamed "loop engineering." Anthropic never titled anything that. The core of all of it is one cycle: gather context, take action, verify work, repeat — and the guidance on simplicity, tool design, and evaluation makes each arrow in that loop reliable. Nothing in any of it promises income.
What holds up
- Anthropic's agent guidance is real, public, and genuinely excellent
- The principles named (simplicity, tool design, evaluation) do appear in the essays
- The agent loop — gather, act, verify, repeat — is the actual documented model
What doesn't
- "$2.3M" appears nowhere; the documents are free blog posts
- "4 years of development, 1000+ hours of lectures" — fabricated dressing
- No single "loop manual" exists; it's separate essays, synthesized by the community
- "An agent that earns" — the docs teach reliability, not revenue
The catch
Fake scarcity is the product. Pricing free documentation at $2.3M manufactures a reason to follow, comment, and enter a funnel. The knowledge was never gated — the engagement is the paywall.
- Effort
- 20 min to read the real essays; the card took less
- Cost
- $0 — every source is a public blog post
- Stack
- Anthropic engineering blog. That's it.
- Confidence
- High
- Posted by
- an AI-content account, ~2.5k followers
PARTLY TRUE "Meet Runneth: the AI brain for the 100X marketer" — a whole marketing team on one always-on machine
The claimA shared AI brain on its own always-on virtual machine that "integrates with every tool in your marketing stack" and runs the whole creative flywheel from Slack — customer research to creative analysis to launch.
The product is real and the architecture is honestly described: every customer gets a genuine dedicated, always-on VM with persistent files, memory, and 24/7 scheduled jobs, from an established creative-analytics vendor with named customers and concrete use cases. What the ad never says: the price. Entry is $750/mo (brands up to $50k/mo ad spend), $1,050/mo above that, extra usage packs $500/mo each, no free trial. And "every tool" is a specific list of ~40 integrations. Strip the wrapper and this is a well-built managed version of an architecture you can assemble yourself with Claude for a tenth of the cost — minus their dataset of millions of tagged ads, which is the real moat and the real reason to pay.
What holds up
- Dedicated single-tenant VM is real — persistent disk, daily backups, 24/7 scheduled jobs
- Established vendor: 2,100+ teams, $14B/yr in ad spend analyzed
- Named customers with specific use cases, not vague testimonials
- Real pricing page, unlimited seats, 30-day money-back guarantee
- Their own FAQ admits you "can get partway" with general AI tools
What doesn't
- "Every tool in your stack" = a fixed list of ~40 integrations
- The ad contains zero pricing — real entry is $750/mo plus $500/mo usage packs
- "100X marketer" is pure puffery, no measurable outcome claimed
- No free trial — money-back guarantee stands in for one
- Their comparison table calls Claude a "single-player chat that can't touch your live data" — stale; MCP connectors do exactly that today
- First-party ad: the poster is the seller
The catch
It's the seller's own ad, and the one number missing is the price: $750–1,050/mo entry, $500/mo per extra usage pack, tiered on your ad spend, no free trial. The architecture being sold — a persistent machine with memory, skills, and Slack + tool integrations — is what Claude Code/Cowork already does for $20–200/mo. What the premium actually buys is their tagged-ads dataset and the managed setup. Price it against that, not against "a whole team."
How to actually do it
- Spending $10k+/mo on paid social and want their creative dataset? Activate the Starter tier ($750/mo, unlimited seats) and treat the 30-day money-back guarantee as your trial — write a list of deliverables it must ship in month one, and refund if it misses.
- Otherwise, build the same shape on Claude: a Claude Pro/Max plan ($20–200/mo) with Cowork or Claude Code on a machine that stays on.
- Connect the same stack via first-party MCP connectors in Claude's settings: Slack, Google Drive, Notion, Canva, and your PM tool. Point-and-click, no code.
- Recreate the "brain": a folder of markdown memory files (brand voice, naming conventions, offers, audience notes) plus reusable skills for repeat workflows — briefs, reports, ad QA. The demo calls this "training"; it is writing things down once.
- The one piece you cannot DIY is their dataset of millions of tagged ads and benchmarks. If that is the draw, that is what the $750/mo is for — price it against one month of a junior analyst, not against the AI.
Split the path: if the tagged-ads dataset matters and your ad spend justifies it, the paid product with its 30-day guarantee is the honest route. Otherwise the same architecture self-assembles on Claude for roughly a tenth of the cost — exact steps below.
- Effort
- Paid route: live same day. DIY route: working brain in an afternoon, genuinely sharp after 2–3 weeks of normal use.
- Cost
- Theirs: $750–1,050+/mo entry + $500/mo per extra usage pack (sold alongside the vendor's core analytics subscription, from $250/mo — their pages are ambiguous on bundling). DIY: $20–200/mo.
- Stack
- Theirs: managed cloud VM + Slack + ~40 integrations + their tagged-ads dataset. DIY: Claude Pro/Max, Cowork or Claude Code, MCP connectors, markdown memory + skills.
- Confidence
- High
- Posted by
- the vendor's own brand page (paid ad, ad-creative analytics SaaS)
PARTLY TRUE "The prompt that writes a digital product in one day"
The claimOne copy-paste prompt — outline an ebook, guide, or template on [topic] with chapter titles, bonuses, and a compelling offer — hands you a sellable digital product, like a "52-page PDF guide," in a single day.
The capability is real and trivially so: AI drafts a book-length guide in an afternoon, and the "secret prompt" is a generic one-liner anyone could write for free. The false load-bearing word is "product." A product implies buyers, and 2026 sales research on the AI book flood shows AI-generated titles make up a large share of marketplace listings but a small share of sales, with lower ratings and weaker revenue — the flood is so thick it drags down earnings for human authors too. Amazon KDP now requires AI-content disclosure and prohibits mass-produced near-identical titles. Writing the guide was never the bottleneck; getting a stranger to want it is. The honest version of this play is excellent, though: a one-day guide is a weak product and a strong lead magnet.
What holds up
- AI genuinely drafts a book-length guide in a day
- The prompt "works" — because any outline prompt works
- Reproduced it live: finished 10-page branded guide in one session
What doesn't
- The "secret prompt" is a generic one-liner, free anywhere
- Outline ≠ product: audience, offer, trust, distribution all skipped
- 2026 data: AI titles = big share of listings, small share of sales
- Amazon KDP requires AI disclosure; bans mass-produced slop
- Prompt-pack funnel at the link in bio
The catch
When writing a book costs nothing, nobody pays for books written for nothing. Generation stopped being the valuable step the moment everyone could do it — the hard part is an audience that trusts you, which no prompt can generate. The account knows this: it doesn't sell books, it sells prompts to people who hope to sell books.
- Effort
- ~1 hour to a finished, designed guide; the audience it needs takes months
- Cost
- $0 with any major AI assistant
- Stack
- Any frontier AI chat + a PDF layout pass — no paid prompt pack required
- Confidence
- High
- Posted by
- a masked "prompt professor" account selling prompt packs via link in bio
PARTLY TRUE Comment "FLOW" for the exact workflow behind "websites of this quality"
The claimComment "FLOW" to receive the exact workflow used to build websites of this quality. The workflow: Direction (goal, audience, concept), Design (structure, interface, typography), Build (development, motion, interactions), Delivery (testing, optimization, handoff).
The rare gate that debunks itself: the "exact workflow" being gated is printed in full, above the gate, in the caption. Direction, Design, Build, Delivery is the standard four-phase agency process — discover, design, develop, deliver — that every studio on earth has run for decades. It's real, it works, and it contains no secret; whatever the DM delivers can only be this same list with a PDF wrapper. The one claim that can't be checked is "websites of this quality": showcase reels in this genre routinely feature award-site footage that isn't the poster's own work, and nothing here proves authorship either way.
What holds up
- The workflow is real — it's the industry-standard 4-phase process, fully disclosed in the caption itself
- Each phase description is accurate, ordinary professional practice
- Not strictly an AI claim — logged because the gate mechanics are the same machinery as the AI funnels in this audit
What doesn't
- The gated "secret" is 100% visible above the gate — the comment buys you nothing you haven't already read
- "Websites of this quality" is unverified — this reel genre is notorious for showcasing other people's award sites
- "Exact workflow" framing manufactures scarcity around the most generic process in the industry
- The gate's real function: comment volume feeds the algorithm and fills a DM pipeline
The catch
This entry is the cleanest specimen of the gate mechanic in the whole audit, because there's no product hiding behind it — the caption gives away the entire "secret" and the gate farms comments anyway. That's the proof that comment gates were never about the content: a comment costs the page nothing, boosts the reel's reach, and opens a DM thread that can be sold to later. If a page ever gates something it hasn't already told you, assume the DM version isn't better — it's just the same thing where the funnel can reach you.
How to actually do it
- Don't comment. You already have the workflow — it's the caption. Here's the operator version of each phase:
- Direction: one page answering who it's for, the one action a visitor takes, and 3 reference sites — this page prevents every expensive revision later
- Design: wireframe before pixels; lock typography and spacing as reusable styles so the build doesn't drift
- Build: develop mobile-first, add motion last — motion is seasoning, not structure; AI tools (Claude, v0) now compress this phase from weeks to evenings
- Delivery: test on a real phone, run PageSpeed, hand off hosting access and an edit guide — handoff quality is what gets referrals
- If you run an agency: notice the tactic itself — gating works even with nothing behind it; the ethical version is gating a genuinely extra asset and saying plainly what it is
Confirmed the gated workflow against the caption it's printed in, and wrote out the operator version of the four phases — the part a DM PDF would actually skip.
- Effort
- 0 minutes — the gated content is already in the caption you read
- Cost
- $0 — the four-phase process is public domain knowledge
- Stack
- Any brief template + wireframe tool + your builder of choice — the process was never the moat; execution is
- Confidence
- High
- Posted by
- a web-design reels account running a comment-to-DM gate on its own process
PARTLY TRUE The "prompt platform": copy a refined prompt, get a working 3D scene "1:1"
The claimLog into the platform, copy a highly refined prompt ("a strict filter that cuts out 99% of useless AI garbage"), and get a working 3D framework or interaction 1:1 — then refine details, saving days of aimless prompt tweaking. Drop "Prompt" for the link; DM "Project" to collaborate.
UPDATED after a follow-up reel leaked the platform's name. The operation behind entries #004 and #007 is Textura Agency, a real digital studio, and the "prompt platform" is GetLayers (getlayers.ai) — a genuine curated library of templates, prompts, and 3D scenes with free samples, a paid Unlimited tier for the prompt library, and a Full Stack tier that includes actual source code, an MCP server, and a Claude Code plugin. That resolves this entry's original paradox exactly as graded: on the prompt tier, "1:1" remains false — LLM output varies per run. On the Full Stack tier, "1:1" is true because you're getting code, not prompts. The product is more honest than its ads: the site itself says "don't ship a layer as-is" and that production-readiness depends on your environment. The reels' anonymous gate-farming and the "strict filter that cuts 99% of AI garbage" framing remain what they were — but there's a real tool behind the curtain.
What holds up
- AI-generated 3D web scenes are real — Claude and peers write working Three.js from a plain description
- "An evening instead of weeks" holds for a working draft scene
- The essay's core (AI collapses the cost of personal projects) is genuinely true — same as this page's entry #004
What doesn't
- "1:1" from a prompt is still mechanically false on the prompt tier — LLM output varies per run
- The 1:1 experience requires the top Full Stack tier, where you're buying source code, not prompts
- "Cuts out 99% of AI garbage" is an unverifiable marketing number
- The reels run faceless with double gates while the company site is public — the ads hide what the product discloses
- Interchangeable philosophy essays around an identical sales block across multiple reels — a template farm
The catch
Rare specimen: the ads are shadier than the product. Textura's own site is upfront — free samples, tiered pricing, commercial licensing, even "don't ship a layer as-is" — while its reels run anonymous, gate comments, and promise 1:1 magic the cheap tier can't deliver. The lesson for the audit: judge the product page, not the reel. And the free-alternative math still stands — Three.js examples and Codrops give you true 1:1 source at $0; GetLayers' value is curation and taste, which is real but optional.
How to actually do it
- For guaranteed-identical results, skip prompts entirely: threejs.org/examples has hundreds of official working scenes with source — copy, run, restyle
- For production-grade showpieces, Codrops (tympanus.net/codrops) publishes free 3D and interaction demos with complete downloadable code
- For custom scenes, describe what you want to Claude like a brief: "a slowly rotating wireframe globe with glowing city markers, dark background, mouse parallax" — expect a working draft in one pass and 3–5 refinement prompts to taste
- In Claude artifacts specifically, ask for Three.js r128 — it's the supported version and renders live in the chat
- If you do want the curated route: GetLayers has free sample prompts and a real Claude Code plugin (Full Stack tier) — evaluate it at getlayers.ai directly instead of DM-ing a gate word
- Budget honestly: one evening to a working scene, a second session to make it client-polished — same arithmetic as this page's own essay
- Never DM a gate word to an anonymous page for access to a platform it won't name in public
Originally graded on mechanism alone; a follow-up reel identified the platform, and inspection of its site confirmed the two-tier resolution predicted here — verdict upgraded to High confidence.
- Effort
- One evening to a working 3D scene draft; a second session for polish
- Cost
- $0 — official examples and demo source are free; Claude Pro ($20/mo) if you want custom scenes drafted
- Stack
- Three.js examples + Codrops demos + Claude for custom scenes — no gated prompt platform required
- Confidence
- High
- Posted by
- a design studio's faceless reels arm — identified via a later reel as Textura Agency, marketing its GetLayers platform
PARTLY TRUE "Anthropic shouldn't have made this free" — the $47B company's "Graph Engineering" playbook
The claimA company doing $47 billion a year wrote down exactly how they run their AI agents, published the numbers, and charged nobody. It's called "Graph Engineering": one lead Claude plans a job and hires a swarm of smaller ones.
More real than it smells — with one invented word doing the selling. The $47B figure checks out: Anthropic disclosed a $47 billion revenue run rate with its May 2026 Series H (run rate, not booked annual revenue — calendar-year projections sit around $20–26B, but the number is theirs). The publication is real too: Anthropic's free engineering post on its multi-agent research system, where a lead agent plans and spawns parallel subagents, with numbers included — 90.2% improvement over a single-agent baseline at roughly 15x the tokens. What Anthropic never did is call any of this "Graph Engineering." That name belongs to a paid AI-builder community whose guide these threads all funnel toward, and this account is one of at least three running the same template this week.
What holds up
- $47B run-rate revenue: really disclosed by Anthropic at its May 2026 Series H
- The publication exists and is free: Anthropic's multi-agent research system engineering post
- The architecture is described accurately: lead agent plans, parallel subagents execute
- The published numbers are real: 90.2% over single-agent baseline, ~15x token cost
What doesn't
- "Graph Engineering" is not Anthropic's term — it's a paid community's rebrand of a free blog post
- Coordinated wave: multiple accounts, same week, same "Anthropic just dropped" template, same destination
- "Shouldn't have made this free" manufactures leak-energy around a post that's been public since 2025
- Run rate ≠ annual revenue — the real number, stretched one notch for the hook
- The omitted cost: the multi-agent pattern burns ~15x tokens, which the thread never mentions
The catch
The tell is the vocabulary. When a thread names a technique something the source never called it, the name is the product — a proprietary-sounding term you can only "learn" from the people who coined it. Anthropic calls this an orchestrator-worker pattern and teaches it for free. The threads call it Graph Engineering and teach it behind a community paywall. Same architecture, one invented word, and the word is where the money is. Credit where due: we went in expecting the $47B to be the fabricated part. It's real. That's why you check.
How to actually do it
- Read the actual source, free: Anthropic's engineering post "How we built our multi-agent research system" — the lead-agent-plus-subagents pattern, with the real numbers and the real tradeoffs
- Then the docs the threads paraphrase: Claude Code's subagents — markdown files in .claude/agents/, each with its own context window, system prompt, and scoped tools
- Try it in any Claude Code project: /agents creates one interactively; a researcher + a writer + a reviewer is the classic starter trio
- Respect the cost line Anthropic itself publishes: multi-agent runs ~15x the tokens of a chat turn — use it where parallel work pays for itself, not for tasks one agent handles fine
- Skip anything charging to teach "Graph Engineering" — the sources above are the curriculum, verbatim and free
Traced the term to its actual coiners, verified the revenue figure and the architecture against Anthropic's own publications, and pointed at the free originals the threads are reselling.
- Effort
- 30 min to read both sources; an evening to wire your first subagent trio into an existing project
- Cost
- $0 for the knowledge — your existing Claude plan covers the practice; expect heavier token use on multi-agent runs
- Stack
- Claude Code subagents (.claude/agents/) + Anthropic's free engineering blog — no paid community required
- Confidence
- High
- Posted by
- an AI-threads account — one of several posting near-identical "Anthropic just dropped" templates the same week
PARTLY TRUE "Claude Code posts to social media for you" — DM POTATO for free access
The claimClaude Code drafts posts, asks for your confirmation, then schedules everything across platforms through the Blotato MCP — no app switching, no logging into 5 platforms. DM "POTATO" on Instagram to get access to Blotato for free.
The capability is real and, for once, honestly described — the workflow in the caption even includes the human approval step. The Blotato MCP server exists (mcp.blotato.com/mcp), connects to Claude Code, Claude Desktop, and claude.ai, and genuinely publishes and schedules to 9 platforms from one prompt, skipping the Meta API entirely. The closing line about MCP giving AI tools to act instead of instruct is accurate. The graded word is "free": Blotato has no free plan — paid tiers run $29 to $499 a month — so the DM gets you a trial, not access. And the caption never mentions the poster founded the company being recommended.
What holds up
- Blotato MCP is real: one endpoint, 9 platforms, publish + schedule from a Claude prompt
- Works in Claude Code, Claude Desktop, and claude.ai as a connector — no Meta API setup
- The caption honestly includes the confirmation-before-posting step
- "This is what MCP does" — the one-line MCP explanation is actually correct
What doesn't
- "Free access" — Blotato has no free plan; $29/mo minimum, so the DM delivers a trial
- Undisclosed in the caption: the poster founded the company the reel recommends
- Adds a paid Claude plan on top — two subscriptions behind a "free" hook
- Instagram publishing requires a Business or Creator account, not a personal profile
- The DM-"POTATO" gate is the same comment-to-funnel machinery as her other reels
The catch
Fourth claim we've audited from this funnel, and the clearest view of how it works: the education is real and generous, the workflows are honestly described, and every road ends at her own $29-plus-per-month product wearing the word "free." To her credit, on her company blog she discloses she's the founder. The reel caption doesn't. That's the tell — the platform where disclosure is legally expected gets it; the platform where the selling happens skips it.
How to actually do it
- Decide the loop you want: drafts-for-approval only (free) or approved-drafts-that-publish-themselves (paid layer)
- Free route: Buffer's hosted MCP (mcp.buffer.com/mcp) on Buffer's free plan — solid for 3 channels and basic queue scheduling
- Paid route: Blotato Starter at $29/mo is legitimately the strongest publish layer if you post video across TikTok, YouTube, and Instagram — one MCP, 9 platforms, no Meta developer-account pain. Sign up at blotato.com directly; the POTATO DM adds nothing but a funnel entry
- Connect in Claude Code with the /mcp command per the setup docs at help.blotato.com/api/mcp, then authenticate in the browser
- Keep the approval gate: prompt Claude to always show drafts and wait for an explicit yes before any publish call — never run it fire-and-forget
- Convert Instagram to a Business or Creator account first if it's still personal — Meta blocks API publishing to personal profiles
Verified the MCP against its live docs and priced the honest version — including when the paid tool is genuinely the right call versus the free route.
- Effort
- 15–30 min to connect the MCP and publish a first scheduled post
- Cost
- $20/mo Claude plan + $0 (Buffer free) or $29/mo (Blotato Starter) for the publish layer
- Stack
- Claude Code + Blotato MCP (or Buffer MCP free) — the publish layer for a draft-and-approve content system
- Confidence
- High
- Posted by
- the same AI educator — this reel promotes the auto-posting SaaS she founded
PARTLY TRUE Drop "Prompt" and get the prompts behind "these websites"
The claimA reel showcases websites supposedly built with AI prompts. Comment "Prompt" and the page DMs you the link to those prompts. The caption's pitch: AI has made an attempt so cheap — one evening — that "I can't do that" is no longer a valid excuse.
Split verdict on a split post. The philosophy in the caption is the truest thing on this page: prompt-to-website tools genuinely do collapse the cost of trying — a competent person with Claude, v0, or Lovable really can stand up an impressive site in an evening. What can't be verified is the load-bearing implication: that the specific prompts behind the gate produce the specific websites in the video. Reels in this format routinely show cherry-picked takes or, worse, showcase sites that were never prompt-generated at all. And the caption's cut-off last line — "that's the whole idea behind our project" — tells you what the reel actually is: an ad with a philosophy essay stapled to the front.
What holds up
- Prompt-to-website is a real, mature capability: Claude, v0, Lovable, Bolt all deliver
- "The price of an attempt is one evening" is genuinely true — and worth internalizing
- The caption is honest that this promotes a project, even if it never names it
What doesn't
- No evidence the gated prompts produced the sites shown — this format is cherry-pick central
- Anonymous page, unnamed "project" — you're DM-ing your engagement to a funnel with no face
- Comment gates exist to train the algorithm and harvest DMs, not to share knowledge
- One "master prompt" doesn't replicate a polished site — real results take iteration passes
- Whatever the link sells arrives only after you're inside the DM automation
The catch
The philosophy is the bait and it's genuinely good bait — cheap attempts are real. But notice what the gate is for: a comment costs the page nothing and earns them reach, your DM thread, and a warm lead for an unnamed "project." If the prompts were the point, they'd be a link in the caption. Verdict would move to High confidence with a screenshot of the websites shown — send one in and we'll re-open it.
How to actually do it
- Skip the gate. Open Claude (claude.ai, Pro $20/mo) and describe the site like a creative brief, not a wish: business name, who it's for, the one action a visitor should take, 3 sections you want, and 2 example sites whose style you like
- Ask for a single-file HTML landing page first — one file is easy to preview, host, and iterate on
- Iterate in plain English 3–5 times: "darker background", "make the headline bigger", "add a testimonial section" — iteration, not the first prompt, is where the polish comes from
- Free hosting: drag the HTML file into Netlify Drop (app.netlify.com/drop) — live URL in under a minute, $0
- If you want the heavier route: v0.dev or lovable.dev generate full React sites from the same kind of brief — free tiers exist, paid from ~$20/mo
- Total honest budget: one evening for a real draft, a second evening to make it good
Graded the checkable claims, flagged the unverifiable one, and handed over the ungated version: how to get a real website out of one evening with AI, no comment required.
- Effort
- One evening to a live draft site; a second session of iteration to make it client-worthy
- Cost
- $0–20/mo — Claude Pro or a builder's free tier, plus free Netlify hosting
- Stack
- Claude (or v0/Lovable/Bolt) + Netlify Drop for hosting — no gated prompt pack required
- Confidence
- Medium
- Posted by
- a faceless AI-motivation page running a comment-to-DM funnel for an unnamed "project"
PARTLY TRUE "Make 1 week of content in 1 minute" — topic in, 7 finished posts out
The claimType a topic into Claude and get one week of content — 7 finished posts from proven templates — in one minute. Comment "CLAUDE" for the social media automation.
The system behind this is real and well documented — a Claude Code setup with skills, brand-voice files, and quality-gate hooks that drafts a week of platform-adapted posts from one sentence. But the poster's own free tutorial for this exact build says to set aside 90 minutes to construct it, and their own workflow includes a human approval step — they state they personally review every piece before it publishes. So "1 minute" is the run time of a machine that takes an afternoon to build and still needs you as editor. The "finished" in "7 finished posts" is the most oversold word in the post.
What holds up
- The build is real: Claude Code skills + brand voice + quality-gate hooks, fully documented
- One-sentence-in, week-of-drafts-out genuinely works once the system exists
- The complete tutorial with exact prompts is free on the poster's blog — no email gate
- Poster is honest elsewhere about reviewing every piece before it goes live
What doesn't
- "1 minute" excludes the ~90-minute build the poster's own tutorial prescribes
- "Finished posts" skips the human review step baked into the poster's own workflow
- The publish pipeline runs through the poster's own SaaS — $29–499/mo, no free plan
- Needs a paid Claude plan on top — the post prices the whole thing at zero
- Brand voice quality depends on feeding it your real writing samples — garbage in, generic out
The catch
Follow the funnel and the "free" one-minute system lands on two subscriptions: a paid Claude plan and the poster's own auto-posting SaaS at $29/mo minimum, which is what "comment CLAUDE for social media automation" is actually selling. Second claim we've audited from this same funnel — the pattern holds: real capability, free education that's genuinely good, and a price tag that only appears after you're three steps in.
How to actually do it
- Install Claude Code (claude.com/claude-code) — a Pro plan ($20/mo) covers a solo content workload
- Follow the free tutorial at sabrina.dev/p/claude-code-full-course-for-beginners — budget 90 minutes, not 1
- When it asks for brand voice samples, paste 5–10 of your real posts — this single step decides whether the output sounds like you or like AI slop
- Keep the quality-gate hook step — it's the part that catches em-dash-riddled generic output before you ever see it
- For scheduling, skip the $29/mo SaaS: Buffer runs a free hosted MCP server at mcp.buffer.com/mcp with published Claude Code setup guides, and its free plan handles basic queue scheduling
- Run it as drafts-for-approval, not fire-and-forget — the creator herself reviews every post, and that's the honest version of this system
Confirmed the build against the poster's own free 90-minute tutorial, priced the hidden subscriptions, and handed over the setup — including a free scheduling route that skips the paid SaaS entirely.
- Effort
- 90 min to 2 hrs to build; after that, ~1 min to draft a week + your review time per post
- Cost
- $20/mo Claude plan; $0 scheduling via Buffer free + MCP, or $29+/mo for the funnel's SaaS
- Stack
- Claude Code (skills + hooks + brand voice files) + Buffer MCP or Blotato API for publishing
- Confidence
- High
- Posted by
- an AI educator with a six-figure following who founded a paid auto-posting tool
PARTLY TRUE "Claude AI makes 1,000 videos for near $0" — no budget, no team, no software
The claimComment "ANIMATION" for 5 free master prompts: Claude AI makes 1,000 videos for near $0 — no budget, no team, no expensive software. You set up Claude Code with a pre-built skill and prompt your way to finished animated videos.
The core is real and genuinely good. Remotion — an open-source video-as-code framework — ships official Claude Code agent skills, so one prompt becomes an animated, fully editable video rendered on your own machine with no per-video fees and no credit system. Independently reproduced: two prompts produced a decent branded 30-second video in about 15 minutes. The hype is the arithmetic. "Near $0" quietly excludes the paid Claude plan doing all the work — one short video reportedly burned ~15,000 premium tokens — and "1,000 videos" is a spreadsheet extrapolation nobody has demoed. Real capability, funnel-shaped framing.
What holds up
- Remotion ships official Claude Code skills: npx skills add remotion-dev/skills
- Renders locally — no credits, no per-video fees; video 1 costs the same as video 100
- Independent reproduction: 2 prompts, ~15 minutes, usable branded 30-second video
- Remotion's free license allows commercial use for individuals and companies up to 3 people
What doesn't
- "Near $0" hides the engine: a paid Claude plan, with heavy token burn per video build
- "1,000 videos" shown nowhere — every usable clip still needs iteration prompts
- First drafts are rough; reproducer's own words: "not beautiful, but pretty damn good"
- For-profit companies with 4+ people need a Company License — $25/seat, $100/mo minimum
- The comment-gated "5 master prompts" are already published free on the poster's blog
The catch
The comment-"ANIMATION" gate is the top of a funnel: DM, then newsletter, then the poster's own paid auto-posting SaaS that this workflow conveniently plugs into. And the gated prize — the 5 master prompts — is already sitting on their public blog, free, no comment required. The real cost they never say out loud: your Claude subscription is the video engine, and at any real volume the token burn is the bill.
How to actually do it
- Install Claude Code if you don't have it (claude.com/claude-code) — any paid Claude plan works; Pro ($20/mo) is enough to start
- In your terminal run: npx skills add remotion-dev/skills — this installs Remotion's official agent skills
- Run: claude — then prompt: "Use the Remotion best practices skill. Create an educational explainer video (1080x1920, 30fps, 30 seconds) that teaches '<your topic>'. Keep all text 150px from the top and 170px from the bottom."
- Skip the comment gate — the 5 ready-to-paste prompts are free at sabrina.dev/p/5-insane-claude-code-video-prompts (explainer, product demo, testimonial showcase, avatar edit, data viz)
- Preview in Remotion Studio (Claude opens it at a localhost URL) and iterate in plain English: "slow the typing", "change accent color", "make the bars wider"
- When it looks right, tell Claude to render the MP4 — output lands in the project folder, ready to post
Verified the workflow against Remotion's official skills, licensing, and independent reproductions, then handed over the exact no-funnel setup — including where the "gated" prompts live for free.
- Effort
- 20–30 min to first video; budget 3–5 iteration prompts per clip you'd actually publish
- Cost
- $0 software; your existing Claude plan ($20–200/mo) is the real engine — token burn is the cost at volume
- Stack
- Claude Code + Remotion agent skills (free for ≤3-person companies) + Node.js
- Confidence
- High
- Posted by
- an AI educator with a six-figure following who founded a paid auto-posting tool
HOLDS UP One free command and "your AI never forgets anything again"
The claimA free video series takes you "zero to your own Jarvis" in about ten minutes: one command, no coding knowledge ever, and you end up with an AI agent with a name and a memory that never forgets anything again.
One of the honest ones. The "AI Memory Vault" is a real, free, open-source system that turns an Obsidian vault of plain markdown files into persistent memory for Claude Code — a legitimate, widely replicated pattern (files on disk instead of a vector database). The repo is live (295 stars, v3.2, July 2026), the series is free, and there's no email gate. The overselling is at the edges: "never forgets anything again" really means "notes are saved to disk" — recall still depends on the agent reading the right notes, and the ten-minute install ignores real-world friction.
What holds up
- Repo is real, free, no email gate — 295 stars, v3.2 released July 2026
- Markdown-as-memory is a proven pattern used far beyond this creator
- Costs disclosed up front: paid Claude plan required, everything else free
- "No coding" holds — Claude runs the build script and interviews you
What doesn't
- "Never forgets" oversells file storage — recall depends on priming and vault upkeep
- "Ten minutes" skips Windows PATH friction, login, and Obsidian setup — plan an hour
- CC BY-NC-SA license: noncommercial — repackaging it for paying clients needs permission
The catch
The funnel is real but soft: free vault → newsletter → coaching and done-for-you services. The giveaway is the ad — it just happens to be a good one. The line nobody reads is the license: CC BY-NC-SA means you can use it yourself for free, but you can't quietly resell it inside client work.
How to actually do it
- Install Claude Code: in PowerShell run irm https://claude.ai/install.ps1 | iex — then add %USERPROFILE%\.local\bin to your user PATH and reopen the terminal
- Install Obsidian (free, obsidian.md) and let it create a vault folder
- Download ai-memory-vault.md from the ai-memory-vault repo on GitHub into your project folder
- In that folder run claude and say: "Read ai-memory-vault.md and build my memory vault." It interviews you and generates CLAUDE.md, VAULT-INDEX.md, and the daily-note template
- Keep CLAUDE.md in the project folder, the vault in Obsidian, and start every session from that folder so the agent boots with its memory
Verified the method against the repo and independent implementations of the same pattern, flagged the noncommercial license, and handed over exact finish-line install steps.
- Effort
- 45–60 min first time (not 10), then a few minutes of vault upkeep per day
- Cost
- $0 on top of a paid Claude plan (Pro $20/mo and up)
- Stack
- Claude Code (native install), Obsidian (free), the ai-memory-vault GitHub repo (free, CC BY-NC-SA)
- Confidence
- High
- Posted by
- an AI educator with a free-content funnel and a ~1,800-member Discord
We test hype for free. We build the real thing for a living.
Thirty minutes, no pitch — and you'll leave with something useful either way.
Book a call with Todd or start with the free Business Checkup →Nothing in that category yet.
Want this pointed at your business?
The audit is free and it stays free. If you'd rather not sort the real from the noise on your own time, start with a Business Checkup — we look under the hood and you keep the findings whether you hire us or not.