Don't Package. Fork.
A pi extension is an npm package that runs code inside your coding agent’s process. When the LLM calls one of its tools, that tool is a function with the same permissions as your agent. Whatever yo...
Circuit Break
Your Notes Are the Wrong Shape for AI
I have 800+ notes in an Obsidian vault. An AI agent searches them dozens of times a day. Some notes return perfect results. Others are useless. The difference isn’t content quality. It’s shape. No...
Your Agent Needs Three Kinds of Memory, Not One
I wrote about persistent memory for AI agents last week. Facts and lessons, extracted from sessions, injected into future ones. 5,000+ people installed it. It works. But it’s one layer of a three-...
Lie to Your Agent (It'll Write Better Code)
“Write tests for this function” produces mediocre tests. “This function has a bug — find it” produces tests that actually catch things. Same agent. Same code. Different framing. Wildly different r...
Turning a Coding Agent into a Personal Assistant
I used to run OpenClaw as my personal assistant. It read my iMessages and extracted tasks into Things 3. It checked Gmail hourly and appended summaries to my Obsidian daily notes. It sent me a morn...
I Built a Native iOS App in Three Days with an AI Agent
I don’t know Swift. I’ve never opened Xcode with the intent to ship something. Last week I had a web app, Pi Dashboard, a browser-based UI for the pi coding agent. By Friday I had a native iOS app ...
Your Agent Forgets Everything Between Sessions
Your AI coding agent is stateless. Every session starts from zero. It doesn’t know you prefer conventional commits. It doesn’t remember that --force-push broke prod last Tuesday. It doesn’t know yo...
Stop Putting Behavioral Rules in Your AI Agent's Prompt
Every AI coding agent has rules: “never force push,” “use conventional commits,” “don’t run dev servers that block the agent.” Most people put these in the system prompt. This works most of the ti...