Status
Two things can be wrong: the platform, and the freshness of the data. The first is Cloudflare’s status page. The second is counted here, live.
Is the data current?
Counted from the database when this page was requested, on 2026-09-06.
| Measure | Value |
|---|---|
| Records linked | 1,734 |
| Sources connected | 2 of 3 |
| Active records verified within 30 days | 1,160 of 1,160 |
| Freshness coverage | 100% |
Both counts are shown so the division is checkable. Where there is nothing to divide the figure reads so rather than printing 0% or 100%, because zero active records is not zero per cent coverage.
The freshness page carries the same figure over time, and data sources breaks it down per feed with the number of days each rate was computed over.
Is the site up?
Live platform status is at cloudflarestatus.com. Workers and D1 are the two components that affect this site; an incident on either is an incident here.
If the page you are reading rendered at all, the Worker and the database both answered — the figures above were queried to build it.
When a source is down
A grantmaker’s portal being unreachable is not an outage here and it is not a programme ending, and the difference is handled rather than guessed at.
A record is not moved to closed on a single failed check: three consecutive failures across at least 48 hours are required. Where an entire source is failing, a circuit breaker trips and no record’s status changes at all — a broken feed is evidence about the feed, not about every programme on it. Per-feed crawl history is published on data sources.
Incident history
There is no incident log on this page, because there is no incident log. Publishing an empty one would imply a monitoring and disclosure process that does not exist yet, which is a worse claim than the absence.
What is published is every deploy, dated, in the changelog.