Latest articles
Stop Asking AI to Fix the Whole Problem: The Zwicky Box Method
Ask a model to fix a problem and you get the answer next door. Take the problem apart first and you get 324. Here's which half the model can do.
Eleven Deep Research Tools That Live Inside Your Coding Agent, and the One I Forked
I forked somebody else's research pipeline, spent 45 commits fixing it, and then wrote a page telling people when to use one of eleven other tools instead. H...
Constraints Breed Creativity (Until You Hand Them to AI)
Ten minutes, keep your hand moving, don't cross out. The most recommended creativity advice in the world is an arbitrary rule. Here's what the rule is actual...
Forced Connections: One Way to Get Original Ideas Out of AI
Ask a model for ideas and you get the listicle. Hand it a fire extinguisher and a rule about how to use it and you get something else.
The Subagents Guide I Wish I'd Had
The other half of the skills guide: what subagents are, how Claude Code actually uses them, and what a year of running two of them on real projects taught me...
You're Using AI Like a Vending Machine
A one-sentence prompt returns the most average possible answer, and that isn't a bug. It's the entire design. Here's what to do instead.
I Got Tired of AI Memory Hype, So I Built a Context Lake
Everyone's bolting memory onto their AI agents. A Context Lake flips it. One brain, outside the agents, that they all read from. Here's why it finally clicked.
The Agent Skills Guide I Wish I'd Had
A working developer's guide to agent skills: what they are, how Claude Code actually loads them, and the quirks for every other coding agent worth using.
I Built an Obsidian OCR Plugin for My Notebooks, Then Started Talking to OpenClaw Instead
I built an Obsidian OCR plugin to read my handwritten notebooks, then stopped using it. The real fix was dictating my notes to an AI agent that keeps me in t...
The Autoresearch Ecosystem - How One Repo Spawned 9 Different Types of AI Projects
Karpathy dropped autoresearch in March and the ecosystem exploded overnight. Here's every category of fork and spinoff, and what each type actually does diff...
My AI Agent Kept Making Shit Up (And Other Lessons From Running OpenClaw)
I set up a self-hosted AI agent to run daily reports autonomously. It hallucinated fake news for two days before I noticed.
Microsoft APM - Managing AI Context Like a Dependency Problem
I went from syncing AI skills across devices to building a full declarative infrastructure for agent context. Here's how APM makes skills easier to use in mo...