Following up on the n8n automation service I posted here about — a second, smaller line I built alongside it.
HyperFrames is an open-source framework (from HeyGen) that renders video from HTML/CSS/GSAP instead of a timeline editor. There's a free Claude Code skill that teaches an AI agent to use it. I almost packaged that skill itself for sale — then realized that's just reselling someone else's open-source work, not building a product.
What I shipped instead: an original guide teaching the actual technique — the mental model behind deterministic, seekable video rendering, plus six complete worked recipes with real code checked against the framework's contract, not simplified examples that break the moment you render them for real. Same legal/business logic as a paid Docker or React course: the tool is free, the teaching around it is legitimately a product.
$19 on Gumroad: https://sidheart.gumroad.com/l/lyudd
Where I am on both lines: n8n automation still has zero real clients (first Upwork proposal is out, no reply yet). This guide is brand new, zero sales, just published. Posting the real numbers as I go, not pretending either one has traction yet.
The distinction between selling the skill and selling the underlying knowledge is solid. Now I’d watch whether people actually pay $19 when the framework and skill are free—that’s the real validation of the guide itself.
"Fair challenge, and honestly the right one to ask. My read: it's not really validation either way from a couple thoughtful comments (self-selected, small sample, as you flagged on the other thread). If it sells at all, that tells me people pay for compressed time/trust in the explanation, same logic as paid courses on free frameworks. If it doesn't, that's real signal the teaching itself isn't valuable enough yet, not a marketing problem to paper over. I'll post the actual numbers here as they come in, same as the n8n line."
Makes sense. The actual sales numbers should make that distinction much clearer.