9 posts in this category
A family-history app for an 81-year-old user, looked after by a headless Claude Code agent that answers his comments and ships fixes every hour. The rules that make that safe, and the bug it found on his first morning.
Read more →We zipped our Claude Code setup into a 37 KB starter kit a friend can install in half an hour. Sorting the files into portable, needs-editing and not-portable was the real work.
Read more →Our TV kept buffering while the mesh app insisted everything was fine. Six days of continuous measurement found one failing node, after four wrong theories. Here is the method, which transfers to any home network.
Read more →We generated a full set of branded video-call backgrounds with an AI image model, and learned that consistency comes from one locked master image and add-only edits, not from better prompts.
Read more →I pointed an AI coding agent at a plan and said 'build this project'. Out came a complete multiplayer browser game you can play right now. Here is what a single prompt can really do, and the one thing that makes it work.
Read more →We built a Claude Code skill that audits any site with Google Lighthouse and produces a fix plan with one hard rule: it must never change how the project renders. The pattern of treating your architecture as a machine-checkable invariant is the bit you can steal.
Read more →We extracted 1,451 of our own prompts from Claude Code's session logs and counted what we kept retyping. The repeats became three new skills and a pile of instructions the agent now reads automatically.
Read more →We took a one-off security-audit prompt, tightened it, and turned it into a global Claude Code skill you can run in any project with /security-audit. Copy the code — or the prompt to generate it yourself.
Read more →The best material for a developer blog is already in your terminal — but letting an AI write it up means guaranteeing it can't publish on its own. Here's the draft-only pattern we landed on.
Read more →