June 2026:
2,493 Commits — A New High

June 2026 set a new high. We shipped 2,493 commits across 20 active projects — up ~43% over May — adding an estimated 870,000+ lines of code across 8,520 files. A team of just three engineers, working alongside a fleet of AI agents, sustained an average of 83 commits per day, peaking at 416 in a single day.

The theme of the month wasn't just volume — it was productization. Two years of building production software with Claude Code became Fabric: a platform our first customers now use to build and deploy their own software, end to end. Alongside it, client delivery kept scaling across mortgage CRMs, loan-origination integrations, pricing engines, and voice/SMS automation.

Two years of building production software with Claude Code, turned into a product. Fabric is an end-to-end portal where you describe what you want, watch it get built, preview it live, and ship to production with one button — the same way we build internally. You design and control every detail; we automated all the hard parts.

It's built for real businesses — especially mortgage. Customers are building CRMs, Encompass / ICE integrations, product & pricing on Optimal Blue, voice AI with Feather, and Twilio SMS/email — wired up and running in production, not demos. This month, our first customers went live on the platform, and we started taking it to market.

I've put about 25 hours into it and it's working almost perfectly. Every issue I flag, it's already fixed by the time I check back. I'm going to be using this a lot — for a long time.

“If we're going to take over the mortgage world, it's going to take a lot.”

— John, early Fabric customer (mortgage)

June pushed 10.96 billion tokens through Claude — a new monthly high — and 10.30 billion of them were read from cache (94.3%). Most of that volume is 1M-token context read straight from cache rather than recomputed: large contexts written once, then re-read across sessions. That efficiency is what lets us nearly double output month over month without a proportional jump in cost. Across the trailing twelve months we've now processed 36 billion tokens at ~94% cache efficiency.

Our development process is fundamentally different from traditional software engineering. We don't work from detailed specifications—instead, we co-develop with AI, iterating rapidly through ideas and implementations.

What might traditionally be called "bug fixes" are really iterative refinements—the natural result of building software collaboratively with AI assistants where the first implementation is a starting point, not a final product.

This approach lets us move faster, experiment more freely, and deliver features that truly match user needs rather than predetermined specifications.

• →Fabric — productized the internal multi-agent build platform for external customers
• →One-button preview → production deploys, with per-branch preview environments
• →Authentication and secure multi-tenant preview access hardening
• →Mortgage integrations — CRM, loan origination (Encompass/ICE), and pricing (Optimal Blue)
• →Voice AI and SMS/email automation wiring (Feather, Twilio)
• →Automated code review, PR correction, and release tracking across repositories
• →Continued client delivery — dashboards, admin tooling, and data sync

• ✓Row-level security hardening across multi-tenant tables
• ✓Reverse-proxy and cross-domain auth fixes so live previews sign in cleanly
• ✓UI/UX refinements across chat, dashboards, and admin surfaces
• ✓Cross-repo synchronization and submodule pin automation
• ✓Schema migrations and data-layer hardening

Files modified this month, grouped by area of the stack:

Activity spread across 20 active repositories out of 203 total in our organization. Ranked by commit volume. Projects are anonymized — the description says what kind of product it is, never whose:

June turned two years of internal practice into a product other businesses can use. Heading into summer, our focus areas include:

• Onboarding more customers onto Fabric and taking it to market
• Deepening mortgage integrations — origination, pricing, voice, and messaging
• Broadening automated review, testing, and quality gates
• Continued infrastructure-as-code and database investment
• Sustaining velocity while tightening cost efficiency at scale