AI-Reviewed Platform Readiness Gate

60 / 60 completed runs
3,000 expected prompts
3,000 eval_run_items
3,000 Lucia responses
3,000 reviews

ready | 60 | 3000 | 3000 | 3000 | 3000

select
  r.status,
  count(*) as runs,
  sum(r.prompt_count) as expected_prompts,
  sum(coalesce(metrics.item_count, 0)) as items,
  sum(coalesce(metrics.response_count, 0)) as responses,
  sum(coalesce(metrics.review_count, 0)) as reviews
from public.eval_runs r
left join lateral (
  select
    count(distinct i.id) as item_count,
    count(distinct i.id) filter (
      where i.lucia_response is not null
        and length(trim(i.lucia_response)) > 0
    ) as response_count,
    count(distinct rev.id) as review_count
  from public.eval_run_items i
  left join public.eval_item_reviews rev
    on rev.run_item_id = i.id
   and rev.reviewer_id = 'user_3EBoNt0THcLkf599Z5bK2UwNHLV'
  where i.run_id = r.id
) metrics on true
where r.created_by = 'user_3EBoNt0THcLkf599Z5bK2UwNHLV'
group by r.status
order by r.status;

ready | 60 | 3000 | 3000 | 3000 | 3000

AI-reviewed platform readiness is not human Lucia-quality approval.