Transparency
The full audit trail.
Trust is earned by showing the work. Below is how our data sources are performing, how fresh the data is, and where independent feeds agree or diverge.
Last cron run: 3h ago
Data freshness
How recent are the underlying observations? Fresh means within 2 years, aging 2–3 years, stale 3–5 years, very stale beyond.
Source uptime
Each data adapter, sorted by 30-day uptime. Failures are surfaced, not hidden.
| Source | 30-day uptime | Runs | Status | Last success |
|---|---|---|---|---|
| CDC/HUD/NICS (reference) | 100.0% | 30 | OK | 4h ago |
| FRED | 100.0% | 30 | OK | 4h ago |
| v2:oecdHousing | 100.0% | 30 | OK | 3h ago |
| AI Index (Stanford) | 100.0% | 30 | OK | 4h ago |
| O*NET | 100.0% | 30 | OK | 4h ago |
| OECD (reference) | 100.0% | 30 | OK | 4h ago |
| v2:eurostat | 100.0% | 30 | OK | 3h ago |
| v2:referenceSeed | 100.0% | 30 | OK | 3h ago |
| BLS | 93.3% | 30 | Failed | 1d ago |
| v2:worldBank | 93.3% | 30 | OK | 3h ago |
| v2:imf | 86.7% | 30 | Failed | 4d ago |
| World Bank | 86.7% | 30 | Failed | 1d ago |
| ACLED | 0.0% | 30 | Degraded | — |
| FBI | 0.0% | 30 | Degraded | — |
How we earn your trust
Every number is sourced
Each indicator value links back to the publishing organization, the reference date, and the raw query we used. No black boxes.
Cross-source validation
Whenever two independent sources publish the same indicator, we display both. If they disagree by more than a set threshold, we flag it and surface the spread.
Failure is public
When an adapter fails, the failure shows up on this page within minutes. We do not hide outages or fabricate continuity.