570,000 Lines of LLM Code Compiled Fine. It Was 20,171x Slower Than SQLite.
Someone benchmarked an LLM-written Rust reimplementation of SQLite. The gap between code that looks right and code that is right turned out to be five orders of magnitude.
← Back Get in touch
Blog
Simple thoughts on building, designing, and shipping.
⌘K
Solo Founder, Zero Employees, $2M ARR: The Agent Stack Making It Real
Four projects shipped in the last two months show what happens when AI agents handle not just coding but earning, orchestrating, and running entire companies.
Create Three Spec Files Before Giving Work to Claude Code or Codex
After a year of agent-assisted development, I found that structured spec files fixed the inconsistency problem better than any prompt technique.
How Codex Solves the Compaction Problem Differently
I reverse-engineered how Codex handles context overflow compared to Claude Code. The answer involves AES encryption, session handover patterns, and KV cache tricks.
Shopify's CEO Built a Search Engine That Revives 700 Claude Code Sessions in One Second
Shopify CEO Tobias built QMD, an open-source search engine. Connect it to Claude Code and every session gets persistent memory.
4 Tool Design Principles Claude Code Learned After 3 Rebuilds
Anthropic's Claude Code team rebuilt their tools three times. Fewer tools made the AI perform better. Here are four hard-won design principles.
Claude Code Sub-Agents Save 25x Tokens in Your Main Session
Your AI isn't getting dumber. Your main session is overloaded. Sub-agents keep it lean and accurate for over an hour.
Claude Code Just Reset Everyone's Weekly Limits to Zero — Here's What Happened
A race condition between Auto Memory and context compaction in Claude Code v2.1.59–v2.1.61 broke prompt caching and corrupted sessions. Anthropic reset all weekly limits as compensation.
Two Tools Every Claude Code User Needs - Agentation and Readout
Agentation gives AI agents pixel-perfect visual feedback via CSS selectors. Readout replays Claude Code sessions like video. Together they eliminate the two biggest friction points in AI-assisted frontend development.
I Ran a 12-Hour Agent Hackathon and Now I Understand Why Stripe Ditched localhost
After building a product with agents overnight, I finally get why Stripe Minions and Ramp Inspect both chose cloud-isolated environments over running everything locally.
Get In Touch
Have a project in mind or just want to chat?
I'd love to hear from you.
I'm always open to a conversation.