Playbooks
The thing to send when someone asks "how do you do that?"
Undated and maintained, with a changelog at the foot of each. This is where the depth lives; the notes stay high level.
Filtering happens on the page: pieces marked "for both" appear under either audience. Prefer a subject to an audience? Browse by topic →
For AI-codersRunning a team or several apps
A continuous security pass for small teams
Scheduled static scanning, database advisor sweeps, a bill of materials matched nightly against vulnerability data, findings routed into tickets with a draft fix. What it catches, what it does not, and the order to build it in.
For companies
An AI policy in one page
A template for the rules part of an AI plan, written to be followed rather than filed. Fill in the named tools and the named person and it is done.
For companies
Before you hire someone to build with AI
The questions to ask an AI-assisted developer or agency before work starts, what a good answer sounds like, and the red flags. About the people building your software, not the AI product you might buy.
For companies
Before your staff build with AI
The fifteen-minute conversation to have before an employee builds something with an AI coding tool: nine questions for the manager, the matching line the builder checks first, and three tiers so saying yes stays cheap.
For companies
Finding where AI actually pays off in a small business
A two-week method for working out where software or AI would save real time or money in a business of ten to two hundred people: map the workflows, score by payoff and effort, check readiness, write it up so the owner can act.
For bothRunning a team or several apps
Moving a business off a no-code platform without losing anything
The teardown-first method for replacing a no-code app that has hit its ceiling: inventory every screen and automation, treat the product as the spec, sort mechanisms into decisions and workarounds, rebuild the flows the team knows, run both side by side.
For AI-codersShipping inside a company
Reviewing code you did not write
The review pass I run on everything an AI coding tool produces: what to read first, the mistakes that show up most, tests as the contract, and how to keep context so the next session does not repeat the last.
For AI-codersRunning a team or several apps
Standards and continuity for a small AI-coding team
The repo files that make AI-written code consistent, how a shared vault survives more than one author, the automated handoff that catches unfinished work, and the drill that proves any of it actually works.
For bothShipping inside a company
The infrastructure around the app
A homework checklist for anyone building an application a business will rely on, and for the leader deciding whether to let them. The app is the easy part; this is everything else.
For companies
What an AI plan actually contains
Five short things with owners, not a fifty-page strategy. The plan for a company of ten to two hundred people, what each part is, how long it takes, and the order to do them in.