Playbooks
A guide explains how something works. A playbook tells you what to do about it, in what order, and how to check that you actually did it. Ten of them, each one written while doing the work for a real business.
How we run a whole agency on one operating system: the desks, the dispatch layer, the receipts every job leaves behind, and the gates that stop work shipping unverified.
The math, page, video, ads and tests behind a call booking video sales letter. A calculator turns your numbers into a daily budget, and tells some agencies plainly that the channel is not worth running for them.
How events businesses show up in Google and AI answers. Seven page types broken into blocks, word and image budgets, verified schema, and the internal linking that ties a season together.
Every band of a town page, top to bottom: the words, characters, photos and links each one needs, measured off four approved pages, and the check that holds every new page to it.
How to give an AI agent a durable, dependency aware work graph that outlives its context window. Harness agnostic, with six health measures you can run against any task graph.
How to build one service page that gets the phone to ring. The order of the page, the copy, the pictures, the code, and the seven things to stop doing on the page you already have.
One main page per service, one page per town your vans go to, all linked back to the middle. Why the hub outranks the pages under it, with every number either sourced or measured here.
What LinkedIn actually does for Google and AI search, checked against what LinkedIn actually serves a crawler rather than what everyone repeats about it.
What Firecrawl does in 2026, where it beats the tools an agency already runs, and the five plays worth running. Every claim tested live or cited from the docs.
A new kind of model that answers a yes or no question in a tenth of a second and tells you how sure it is. What it does, who it is for, what it costs, the nine ways it breaks, and how to start in shadow so it earns every rung.
How to build the page where somebody decides the people coming to their house are real. The facts a stranger can check, the photos, the code, and why most shops should not have a team page at all.
Everything we check before a home service website goes live: links, indexing, schema, what AI tools can read, and speed on a phone, each proven by a check that has been broken on purpose.
How an AI agent finds what Google Search Console says is broken, fixes it at the source, and proves it fixed when Google crawls again, with one person saying yes before anything ships. The skill and script are free to download.