quality · the rubric board
What “good” means here.
Every tool, lesson, and game on Lotus Mind carries the same bar. If something misses it, you can say so right on the item — and we count it.
Six standards
- 01REQUIRED
Named model every answer
AI output always says which model produced it. No mystery meat.
- 02REQUIRED
Sources or it didn’t happen
Study claims link first-party sources; unproven patterns are labelled INFERRED.
- 03REQUIRED
Human review before release
Lessons and legal/health-adjacent copy pass a person before release, with sign-off recorded.
- 04REQUIRED
Feedback that lands
Every game, tool, and lesson has thumbs + a comment box. Weekly triage of the top signals.
- 05REQUIRED
Accessibility a release gate
WCAG 2.2 AA checks per page; blocking issues block the release.
- 06REQUIRED
Teach the errors in public
Post-mortems of our own misses are published in Docs.
Hold an item to the bar.
Slide each standard for any game, tool or lesson you’ve used. The board shows whether it would ship. It’s how we triage — try it on something.
Runs on this page only. Nothing is sent.
Held — accessibility is a release gate.
The loop, in four moves.
- 01
Measure
usage + explicit ratings
- 02
Triage
weekly, human
- 03
Fix
root cause, not paint
- 04
Report
what changed & why
Live status of every surface: Service status · error budget and incident history included.
Lotus Mind Limited · 10 Aderrig Manor, Adamstown, Lucan, Dublin K78 A4T8, Ireland