
One Set of Rules to Support Multiple AI Coding Agents
I didn't want to be locked into one AI coding tool. Here's how I keep a single set of rules, skills, and protocols working across several agents at once.

~/home
Product Leader & AI Builder
Hey, I’m Maxim. I’m a Staff Product Manager at Udemy, leading MarTech product strategy, with over a decade of experience across marketing, technology, and product. I write about product, AI, and what’s changing about how teams build, ship, and achieve outcomes. And I build at the intersection of product and AI.

I didn't want to be locked into one AI coding tool. Here's how I keep a single set of rules, skills, and protocols working across several agents at once.

A different vendor's model reviews what my primary AI built, but only on the changes where the stakes justify it. Here is the cross-vendor review system I run, and what it does not fix.

CRO cycles take weeks across multiple people and often ship on intuition. Here is how I ran one as an agentic workflow with a research step that caught flawed recommendations before they shipped.
This site. Where I write about product, AI, and what's changing about how teams build, ship, and deliver outcomes.
Free strategic product planning platform for PMs. Turns product vision into strategic horizons, quarterly plans, and trackable initiatives.
Exploring the vendor-agnostic orchestration layer above LLM primitives. How to wire skills, hooks, and agent patterns into something production-grade that isn't locked to a single model provider.