Operating principles

Our values tell you what we care about. They don't tell you what to do at 4pm when you're staring at a pull request and two reasonable options. That's the gap these fill. A named rule you can say out loud in a review moves a decision in a way a poster never does, so we gave the hard-won lessons names. When someone says "let's ship the thin slice" or "reversible, so just decide," the shorthand does real work. Here's the list.

The principles

Thin slice. Ship the smallest change that moves things forward, then iterate. A big change is a stack of small ones you haven't split up yet. See ship the smallest useful change.

Boring on purpose. Reach for proven, well-understood tech before the exciting new thing. Novelty is a cost you pay every time something breaks at an inconvenient hour. Spend it where it buys the user something.

If it isn't written down, it doesn't exist. Decisions, processes, and reasons live in the handbook or in the PR, not in someone's memory. Writing it down is how a crew this small keeps working after the person who knew leaves the chat.

Disagree, then commit. Argue hard while the decision is open, then get fully behind the call once it's made, even if it wasn't yours. Re-litigating a settled decision in the hallway is how teams rot.

No turf. Nobody owns a corner of the codebase or a category of decision. If you can improve a thing, improve it; if someone improves yours, say thanks. Guarding territory is slower than just fixing the bug.

Hold your opinions loosely. Have a strong take and be genuinely ready to drop it when someone shows you a better one. Being right matters more than having been right.

Delete happily. The best change is often the one that removes code. Less to run, less to break, less to explain. Take the win and move on; git remembers if you're wrong.

No heroics. A 2am rescue is a symptom, not a badge. We'd rather fix the thing that made the heroics necessary than celebrate the person who stayed up. Sustainable beats impressive.

Assume good intent. Read the generous version of what someone said before you react to the worst one. Most friction is a missing sentence, not a bad actor. Ask before you assume.

Make it reversible. When you can turn a one-way door into a two-way one, do it, and then decide fast. Cheap-to-undo decisions don't deserve slow deliberation. Save the caution for the calls you can't take back.

Good, then out the door. The bar is good and shipped, not flawless and imaginary. Perfect is a way of never finishing. Take the extra day when it's warranted, not the extra month because you're scared to demo it.

Show your work. When you decide something, leave the reasoning behind so the next person understands the "why," not just the "what." A decision without its reasons is a landmine for future you.

Own the whole outcome. "Not my department" isn't a thing here. If you see the problem, it's yours until it's handed off or fixed. The gaps between people's jobs are where things quietly break.

No ceremony. No status meetings to prove work happened, no decks to make a simple point, no ritual for its own sake. If a thing exists to look like progress rather than to make it, kill it.