Questions

Why the corpus is small on purpose, what "reviewed" means here, and why this does not try to out-write Wikipedia.

Why so few articles?

Because the alternative is worse. A generated encyclopedia can produce thousands of pages in a day, and doing so would produce thousands of pages nobody checked — which is both a bad product and a spam-policy problem. Depth first is a decision, not a stage we have not reached yet.

What does "reviewed" mean?

A person read the article against its sources and put their name to it, with a date. That is all it means, and it is more than most generated pages can claim.

Why not just cover everything Wikipedia covers?

Because we would lose, and we should. Wikipedia is better at general reference than we will be. The questions worth answering here are the ones a reference work structurally does not do well: comparisons, causes, and the things people routinely conflate.

What happens when an article is wrong?

It gets corrected, with the date changed and the correction visible. A silent edit is how a reference work loses the only thing that makes it worth reading.

Questions

Can I cite Leumapedia?

Cite the sources listed on the article instead — they are the actual authority, and they are on the page for exactly that reason.

Do you use AI to write these?

For requested articles, yes, grounded in real sources and reviewed by a person before publication. Every article states which it is.