Principles of bettergoals.ai

These principles guide everything on this site — the content we curate, the features we build, and the way we build them. They are community-owned: propose a change by editing this file on GitHub and opening a pull request.

1. Outcomes over outputs

A better goal describes a change in the world — for a customer, a colleague, a citizen — not a list of things to build. If you can complete a goal without anything getting better for anyone, it isn't a goal, it's a task list.

2. Better value, sooner, safer, happier

Every goal should be able to answer four questions: What value? How will we see it sooner? What makes this safe — to attempt, to challenge, and to miss? Who ends up happier?

3. Small enough to learn from

Prefer goals you can test in weeks over ambitions you can only judge at year-end. The point of a goal is to create a feedback loop, not a verdict.

4. Safe to challenge

The best predictor of goal quality is whether the people closest to the work felt safe to say "this is the wrong goal." Psychological safety is a prerequisite, not a nice-to-have.

5. Built in the open

Everything on this site — content, features, even these principles — is contributed, discussed, and endorsed in the open. Ideas live as GitHub issues; changes ship as pull requests anyone can review.

6. Accessible to novices, powerful for experts

A first-time contributor should get from thought to board in two minutes. An advanced contributor should be able to ship a voice-coaching agent. Both belong here.

7. Humans endorse, AI accelerates

AI (Claude Code) builds what the community endorses — it never decides what's worth building. Endorsement, review, and shipping to production are human acts.