I used to think distribution was mostly a messaging problem.

Now I think it is a systems problem.

The market is shifting fast:

• frontier labs are turning models into infrastructure
• open-weight models are attacking the wedge
• agents are changing how builders work
• token cost, context, and distribution are becoming product problems

So I build tools around one question:

What hidden work is already valuable, but not captured yet?

Most people try to create more content.

I’m more interested in capturing the value already buried inside:

• build sessions
• agent traces
• prompt context
• debugging logs
• product decisions
• technical failures
• market observations

Turn messy work into usable systems.

That is the loop.

CacheCatch is a prompt-cache audit tool for AI agents.

It scans local coding-agent sessions and trace data, then shows:

The goal is simple:

find repeated context → detect cache breaks → fix prompt layout

Not generic observability.

Not another dashboard.

A cache report builders can actually act on.

SpielOS is my broader system for turning founder work into distribution.

At the end of a build session, you run:

/post

SpielOS turns the work you just did into:

The flow:

session → context → draft → format → publish

Not “write me a viral post.”

Content from actual work.

Technical founders do not need to become full-time creators.

They need systems that turn their work into public proof.

Every serious builder already has raw material:

• debugging decisions
• failed attempts
• product tradeoffs
• agent conversations
• architecture notes
• shipping logs
• weird technical discoveries

But most of it dies inside terminals, IDE agents, private notes, and forgotten chats.

I build systems to stop that.

The common thread:

Capture the hidden system inside the work.

Every build session already has a story inside it.

The problem is not:

make more content

The problem is:

capture → structure → narrative → distribution

That is the layer I’m building.

I come from marketing strategy.

That means I do not only care whether something works technically.

I care whether it creates leverage.

Can it be understood?

Can it spread?

Can it compound?

Can it create a new behavior?

That is the part I keep obsessing over.

These are the questions behind most of my work:

• Can technical founders distribute without becoming content creators?
• Can AI agents become part of the product workflow, not just coding assistants?
• Can prompt caching become the new prompt engineering?
• Can build sessions become reusable assets?
• Can local tools beat bloated SaaS for serious builders?
• Can AI market shifts be turned into useful founder tools?
• Can strategy games become more political, agentic, and alive?
• Can work become content without turning builders into influencers?