Field notes.

I've spent the last year shipping AI into other people's companies. Here's what actually works in production, what's still hype, and the playbook I use on every new engagement.

After a year of installing AI into other people's companies, these are the five skills I install on almost every engagement. Highest reliability, lowest setup cost, easiest to hand off.

Client reporting is the highest-frequency manual task in any agency-shaped business. Here's the no-code skill I install on every new engagement, and how I keep it running without becoming the bottleneck.

Three engagements in I had three subtly different codebases, none of which talked to each other, and every upstream change broke all three at once. Here's how I switched to a forkable, version-controlled cookbook — and why integrator economics live or die on this decision.

Every integrator gets the same call: 'can you make the AI do X for our specific case?' Here's the workflow I use to take a custom request from a 30-minute conversation to a production-ready skill, without writing a one-off project from scratch every time.

How to turn OpenClaw from an amnesiac chatbot into a working partner — memory, soul file, heartbeat loop, and sub-agents. Paste this into your agent and it will start implementing.