Essays, playbooks, and measured findings on building with AI, running agencies, and how work actually gets done - plus personal stories and the opinions I will defend. Much of it published here first, and all of it open.
A standalone recipe for building a remote MCP server that Claude.ai connects to over OAuth: architecture, Auth0 setup, the security isolation pattern, tool design, deployment, and testing. Hostnames, ports, and credentials are placeholders - swap in your own.
A website can look perfect in your browser and be empty to the machines that increasingly decide who finds you. This is the audit I run on every site: seven areas, each with a test you can do right now, ordered by how much damage failing it causes. Free tools, small fixes, no SEO agency required.
Somewhere right now, a business owner is discovering that their website's hosting is registered to an agency that no longer answers email, and nobody knows the CMS admin password. This playbook is about being the vendor after whom that story is impossible. It costs you almost nothing. It is worth more than most portfolios.
Most LinkedIn profiles are written like résumés: a list of jobs, read from the owner's point of view. The profiles that generate opportunities are written like landing pages, built around one clear message and read from the visitor's point of view. This is the guide I wrote after rebuilding my own, section by section, in the order visitors actually see things.
MCP is the least-explained important thing in AI right now: developers talk about it constantly, buyers hear the acronym and glaze over, and almost nobody has written the version for the person who signs the budget. This is that version.
Every custom development project starts with the same uncomfortable gap. The buyer commits real money to a builder they haven't worked with, for a system that doesn't exist, based on a proposal that is a prediction. There is a better structure - small, paid, credited on continuation, with deliverables that stand on their own.
There has never been a better time to buy a small custom AI tool, and there has never been an easier time to waste money on one. The difference between the buyers who get value and the buyers who get burned is almost never technical knowledge. It is knowing what to ask for. This is the checklist I wish every buyer walked in with.
I write regularly on LinkedIn about what I'm building and learning: agency growth, AI development, product judgment, and the messy reality behind making things work.
Follow on LinkedIn