Writing
Thoughts on development, AI, automation, and building cool stuff.
Anthropic's June 15 programmatic credits split changes how you pay for agent usage. Here's what the numbers mean and why most builders are missing it.
MCP is how domain data becomes AI expertise. Thomson Reuters just proved it by rebuilding CoCounsel on Claude's Agent SDK. Here's what that means.
Everyone's still arguing about whether AI can code. Meanwhile Skills and subagents quietly rewrote how I work — and most developers haven't even tried them yet. Here's what actually changed, with real examples from my own setup.
I went from hitting rate limits every single day on Claude Pro to shipping features at a pace I didn't think was possible solo. Here's what actually changed—and why the gap between AI-native developers and everyone else is becoming impossible to close.
Claude Code shows you what it's doing live — but once a session ends, you can't replay it, diff it, or regression-test it. I built AgentTape to fix that.
How I built a completely automated payment system that sends digital resources instantly when customers pay and follows up on unpaid invoices—all without touching code in production.
