Engineering things
with AI.
I build things — products, tools, systems. AI is how I work. This is where I share what I ship and what I learn.
Selected Work
PiBun
A native desktop GUI for the Pi coding agent — chat, terminal, git, plugins, themes — built with Electrobun, React, and Bun. Open source.
QuoteCraft
A mobile-first quoting app for trade contractors — send a professional branded quote from your phone in under 3 minutes. Expo + Convex, built for the field.
Second Brain
An AI-powered personal knowledge system — dump your thoughts, and AI files, connects, and surfaces them when they matter.
Singapore Legal SEO
Built a multi-country legal data platform — 18,000+ court judgments, 44,000+ entity profiles, AI-generated content — across 4 Commonwealth jurisdictions. Effectively $0/month.
Systems
The .plan/ protocol Four files that replace standups, sprint planning, and handoff meetings when building with AI. Designing an AI agent's soul SOUL.md, HUMAN.md, and TENSIONS.md — three files that turn a generic AI into a thinking partner that actually knows you. Work on the system, not in the system The operating principle behind everything I build — and why most engineers are doing it backwards. Autopilot: unattended AI execution A shell loop that runs an AI agent repeatedly against a plan — each iteration gets a fresh context window, does one thing, verifies the build, and exits. No human needed. Teaching agents to know what they don't know AGENTS.md and CAPABILITY-MAP.md — two files that make AI agents honest about their limits, and resourceful about overcoming them. Atomic Design was made for AI agents A living design system built on Atomic Design Methodology — where the AI agent works backward from pattern specs, composing primitives and tokens it already knows. The methodology finally makes sense when the maintainer never gets tired. Give the agent a browser browser-tool.ts — a thin Playwright CLI that turns an AI agent into a web researcher. No scraping framework, no complex automation. Just launch, screenshot, decide, act. Meta-tasks: when the agent writes its own plan CREATE TASKS — plan items where the agent's job isn't to do work, but to read the world, decide what work should exist, and alter the plan. The plan becomes a program that generates more of itself. Deep modules: structuring code for AI agents, not humans Most codebases are structured for human developers navigating with IDEs. AI agents don't use IDEs — they use tool calls. Every file read costs time. Every cross-file edit risks failure. Here's how I'm restructuring a 33,000-line codebase to be maintained by AI.