Build list — one sitting

Build the machine in 30 minutes

It fits, because almost none of it is building. Insaight is already the store and the research layer. You already have the templates. I write the skills while you click through three logins.

Nothing here has a waiting period any more — no domain to warm, no ID check to clear. Your total keyboard time is about ten minutes; the rest is me typing and you reading.

{{ doneCount }}/8
You= you have to be the one clicking Me= I do it, you watch
{{ t.mark }}
{{ t.time }}

{{ t.title }}

{{ t.who }}

{{ t.body }}

At 0:30 you have a working machine

It scores opportunities against your profile, finds decision makers, researches them, generates a site, fifty ads and a deck each, publishes to companyname.danilovicioso.com, and hands you an approval page. It runs itself nightly from then on.

Both channels are live from day one: email sends automatically from nilovicioso@gmail.com, LinkedIn gets queued and you fire it in eight minutes a day.

Two things that got cheaper

No Meta developer account
The Ad Library is a public web page. Claude Code reads it directly — no app, no API key, no ID upload, no review queue. Slower per company than the API and perfectly adequate at ten a night.
No sending domain, no warm-up
nilovicioso@gmail.com is an aged, real, human-looking inbox — which at 20–30 sends a day lands better than a brand-new domain would after three weeks of warming. Forward replies to your main inbox and it costs nothing.

The one rule that comes with using a Gmail. Google tolerates low-volume personal mail and suspends accounts that look like bulk senders. Stay under about 30 a day, keep every message genuinely different — which your templates already are — and never send the same body twice. Cross that line and you lose the inbox, not just the campaign.

What I'd cut if we run over

In order. Each one is recoverable in a later sitting and none of them stops the first prospect going out.

  • Cron. Run it by hand for the first week. You want to watch what it does anyway.
  • The wildcard DNS. Raw Cloudflare Pages URLs work fine for the first few prospects — uglier, not broken.
  • The deck skill. The site and the ads carry the pitch. Add the deck once you've seen which prospects reply.
  • The approval page. Read the generated folders directly for the first ten. Build the review screen once approving is genuinely tedious.

One thing to decide before we start. Which pool the first run targets — A (hardware, weak funnel) is the safest test because the demo argues for itself, D (baseball) is the most fun and the fastest to a warm reply. Pick one and I'll write the scoring skill against it rather than trying to serve all six on day one.