First-party only — a session id and where you arrived from, so we can tell which pages are worth writing more of. No advertising, no third-party trackers, nothing that follows you off this site.
The whole first track is free and always will be. It covers the things that decide whether an agent is useful on day one: what to tell it about the repo, how to scope a request, when to stop it, and how to read what it produced. If you are new to working with a coding agent, start here and go in order.
9 lessons · about 63 minutes of reading
The job does not become "writing prompts". It becomes specifying, bounding and reviewing — and those three are where all the difficulty moved.
CLAUDE.md, AGENTS.md and their cousins are the highest-leverage thirty minutes you will spend. Here is what belongs in one and what does not.
The difference between a useful agent and a frustrating one is usually the size of the change you asked for. Four techniques for keeping a diff reviewable.
Agent-written code fails differently from human-written code. A four-pass review order that finds the expensive problems first.
The single highest-return habit in agent debugging. Confident wrong theories are the default; a failing test you watched run is the fix.
Approving every command is unusable; approving everything is reckless. A permission policy that is safe and still fast.
A long conversation is not a rich one. What to include, what to leave out, and when to start over.
Knowing where an agent stops paying is a skill, not an admission. Six situations where the handoff costs more than the work.
The whole track applied end to end. Ninety minutes from cloning something unfamiliar to shipping a correct first change.