The whole machine, documented and yours to run.

Because the system ships when it's proven, not when a countdown timer says so. It's being packaged in public right now — the channel documents that work with receipts. The founder pre-order opens before launch, the list hears first, and until then nothing is purchasable. That's deliberate: you should be able to watch the machine work before you're asked for money.

No — with honest scoping. The course teaches you to understand and direct the machine, not to hand-code one: it explains the stack from zero — terminal, Node, API keys, the agent — and shows every stage running for real. On the system tiers (02 and 03) you set up the shipped app by checklist, no code written. On tier 01 alone, building your own version of the machine is your project — the course gives you the blueprint and the disciplines, and your own AI agents do the building.

Tier 01 is the recorded course — being produced now from the running system, delivered at launch. Tier 02 adds the machine: a download of the complete app and its working files as a dated edition snapshot. Tier 03 adds the onboarding call and the 30-day support window, which opens at launch.

The machine ships as an edition, frozen at a date and complete at purchase. There is no server of mine to shut off and no subscription that lapses. Future editions, if any, are optional paid upgrades — no obligation on either side.

The honest caveat: the machine stands on outside services — the model APIs, the voice provider, the renderer under the visuals stage — and those vendors change things on their own clock. A frozen edition keeps working while its outside pieces hold still; when a vendor retires a model or an API, adapting the edition is operator work. The system is plain-text specs and plain API calls precisely so you — or your AI agent — can make that kind of repair.

A Claude Code subscription for the agent, a voice key (ElevenLabs), an image-generation key (Google Gemini or OpenAI, your pick, as the system runs today — the roster can change between editions), model API keys for the fact-check matrix (Anthropic, OpenAI, or Gemini — any mix), and YouTube API access for research and analytics. You pay each provider directly for exactly what you use — no markup, no middleman.

On running costs, the honest answer is receipts, not a marketing number: image generation shows its exact price before you approve it, the system enforces per-video budget caps, and a ledger tracks the spend. The course will show real, calculated cost sheets on screen, so you see what a finished video costs before your first key exists.

No — and that's deliberate. The system hands you a complete publish packet (title, description, tags, thumbnail); you press upload in YouTube Studio. Its only connection to your channel is read-only analytics. Nothing ever posts on your behalf, so no automation mistake can ship to your audience.

The system is plain Node, a local database, and plain-text specs — but honest labeling matters here: it's built and verified daily on Windows, and macOS is not a tested, guaranteed path today. If that changes, it changes on the record.

The system is channel-agnostic by design — your niche, voice, and visual identity live in a channel profile and settings the setup walks you through. It's built for long-form educational and explainer content. Narration can be voiced in the languages your voice provider supports; the drafting and quality tooling is built and battle-tested in English first — non-English drafting works, but is younger.

YouTube's monetization rules target what it calls inauthentic content — in its own words, "mass-produced or repetitive content", videos "made with a template with little to no variation across videos", where the standard it demands is that "the substance of each video should be materially varied". For AI specifically, what it flags is "AI-generated content made with generic templates" made "without adding the creator's original, authentic insights or perspective".

The system is built to honor those rules by design: every video starts from its own research and an angle you approve, the script is drafted fresh and fact-checked under your gate, and nothing is stamped from a template into the next video. But honoring rules is the operator's work, not a property of software — the same tooling pointed at templates and volume would produce exactly what the policy targets, and nothing in the box can stop someone from using it that way. You control the quality and authenticity of what ships; the system's job is to make the diligent path the easy one. And no honest seller can promise a platform's decision about your channel — the policy is public, the discipline is yours.

No income claims — not here, not in the course. What the system demonstrably does is production leverage: one person shipping two quality long-form videos a week, with receipts. What a channel earns depends on things no honest seller can promise.

The statutory right of withdrawal applies to consumers in the EU; the details and the withdrawal form are at the withdrawal page, and the full terms are shown before any purchase.

Because the product is a system, not a membership. A solo operator selling standing obligations stops being a solo operator — and you'd be paying for my time instead of your pipeline. Everything you need to run the machine ships in the box; tier 03 exists for a bounded, honest start, and then it ends.