Data Health

Is every feed actually alive, or are you flying blind on stale data? Every number in this tool is a live query — but a query is only as fresh as the daemon feeding it. Anything red here is a feed that quietly died.

Reading this: green = fresh · amber = going stale (hours old) · red= dead (>24h tables, >7d handles). A dead handle = that account's scraper broke. This is the panel that catches a feed dying before it quietly poisons everything downstream.

The readAll feeds fresh0 tables · 0 handles

Every feed is fresh — downstream signals are reading live data. Data freshness is a prerequisite for signal quality: a stale narrative_ranks table means Foresight is lying, a dead KOL handle means assertions haven't processed recent posts, and a dead convergence_events table means paper trades aren't firing.

✓ no major caveat on the headline read

→ If any table is amber or red, check the daemon log: python scripts/watchdog.py --json — it will show which daemon went stale and respawn it with --restart.

Tables

Per-table row count + how long since the last write.

KOL handle staleness

Per-handle last-post age across the Apify whitelist. A handle dark for >7d = its scrape path likely broke.