The weekly update
Every week the lessons from real work go back into the guide. Each page here is that week written up in full: what happened, why it mattered, and what changed in the guide.
- Check the thing itself This week is about checking the real thing before acting on an assumption about it.
- Say only what actually happened This week is about the gap between what was said and what was really true or really done.
- Do the part that is yours Work kept landing back with the user when the assistant should have carried it.
- The user’s view is the one that counts What the assistant can see is not always what the user is looking at.
- The model of the thing, not the thing The assistant worked from its own picture of things when the real thing was right there.
- The check that reported success A check can run, pass, and never touch the thing it was meant to check. This week is about making sure it does.
- The fact nobody checked This week’s work was built confidently on facts that were assumed rather than checked.
- The loop that never told anyone what it was costing Some systems keep reporting success long after the thing you care about has stopped being true.
- The gate that was only a sentence Intending a check is not enforcing one. A rule that must hold cannot live in prose written by the thing it constrains.
- What’s true after the turn ends This week is about the gap between when the assistant stops looking and when you look.
- The report nobody drains Nothing errored and every report was right. The failure lived only in the repetition.
- The map is not the territory Every lesson this week is a case of trusting a description of the system over the system itself.
- Quiet failures that report success The quiet failures are the ones that report success. Nothing errored, and something was lost every time.
- Failures in disguise Concurrency and defaults fail in disguise. The real cause never shows up in the error.
- Look before you act Most of this week’s mistakes came from acting before looking at what was already known.
- Self-healing background runs A background run can do more than report. It can guard a risky live change and undo it if the change does not take.
- Mine your history for hooks The best source of new hooks is your own session history. Group past sessions by what went wrong, and the top failures are your backlog.
- The other assistant in the room A second AI assistant working beside yours needs territory, not walls. Plus: ship without asking, read the mode before the command, and look before you ask.
- Workflows, skills and hooks Into the guide: scripted multi-agent workflows, skills that fire before the work starts, and a hook fault that only shows with two sessions open.
- Defaults and sources Rewriting one cautious sentence changed more than a pile of exceptions. A few standing defaults about how to act matter most.
- Defensive rules pile up Cautious rules can quietly add up to an assistant that asks before every step. The fix is to rewrite the default, not to add another rule.
- A new baseline The guide was rewritten from the ground up for people who have never used Claude Code. This is the new baseline.