Becoming AI-native is on every company’s roadmap. It’s also on the mind of every person inside one. Neither side has the playbook yet.
First Layer Labs is how I’m exploring this problem: building small tools and trading notes with operators working on the same thing.
I’m learning AI by using it in my own work. As I find aha moments along the way, I share them here, hoping they help you find your own. Best for non-technical employees at companies using Claude, but anyone bringing AI into their work might find something useful.
A hands-on flow that turns a generic Claude into one set up for the work you actually do.
Try it ↗A Claude skill that strips AI tells from drafts while keeping the writer’s voice. One file, two install paths, yours to customize.
Install the skill →A prompt that looks at your last seven days and gives you three specific things AI would have made better.
Read & copy →Inventory every connector you have set up. Surface the three highest-leverage things each one could do for your role today. A good first prompt if you’ve just connected something and want to see what it can do.
Read & copy →A prompt that builds you your own “think bigger” skill, so your AI catches you when you’re scoping an idea too small.
Read & copy →Same idea, built for Claude Cowork. A flow that turns a generic Cowork setup into one that works like a personal OS for your actual job.
In progressAs co-founder of Repeat (acquired in 2024), I spent years as a product-minded founder obsessed with two questions: how to get customers to actually adopt and use a product, and how to build an organization that ran efficiently and moved in one direction.
Both questions are at the center of AI right now. Companies are working on both at the same time: getting their people to actually use these tools, and reshaping the work around them.
First Layer Labs is where I’m applying those learnings to AI.
Either way, I want to hear from you.
Get in touch →