Pipeline & Lineage
One row per collector, 14 days. Computed from core.collection_logs and the trading calendar — not from the worker’s verdict, so it can disagree with it.
Worker verdict: PIPELINE_OK — 24/24 lanes healthy. This page re-derived the same question independently and agrees.
| Collector | 07 | 08 | 09 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | due |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| cboe_indices | sessions 22:30 ET | ||||||||||||||
| fred_ice_oas | weekdays 12:00 ET | ||||||||||||||
| fred_tier_a | weekdays 06:00 ET | ||||||||||||||
| metric_daily | sessions 18:30 ET | ||||||||||||||
| occ_open_interest | weekdays 21:45 ET | ||||||||||||||
| tiingo_prices | sessions 18:05 ET | ||||||||||||||
| uw_borrow_rates | sessions 17:00 ET | ||||||||||||||
| uw_earnings | weekdays 06:30 ET | ||||||||||||||
| uw_economic_calendar | every_day 20:45 ET | ||||||||||||||
| uw_etf_holdings | weekdays 16:30 ET | ||||||||||||||
| uw_etf_profile | weekdays 16:30 ET | ||||||||||||||
| uw_etf_weights | weekdays 16:30 ET | ||||||||||||||
| uw_flow_alerts | sessions 21:15 ET | ||||||||||||||
| uw_interpolated_iv | sessions 17:00 ET | ||||||||||||||
| uw_multileg | sessions 17:15 ET | ||||||||||||||
| uw_news | every_day 20:45 ET | ||||||||||||||
| uw_option_chains | sessions 19:00 ET | ||||||||||||||
| uw_predictions | every_day 20:45 ET | ||||||||||||||
| uw_risk_reversal | sessions 17:10 ET | ||||||||||||||
| uw_variance_risk_premium | sessions 17:05 ET | ||||||||||||||
| xnys_calendar | every_day 03:00 ET |
collected (158)below the row floor, or degraded (see the run's notes) (5)ran, returned nothing!raisednot a session✕DUE AND NEVER RANdue later today (4)not due (78)before this lane existed (49)
Impact — if a source dies, what goes dark
Straight from each source’s declared consumer. GOVERNANCE R5 requires one before a collector is built, which is what makes this column possible at all rather than a diagram somebody has to maintain.
- cboe_indicesRESTATED · export restricted
- core.macro_observations - the regime view's volatility axis (VIX and its percentile, DESIGN.md:147) and Page 5's VIX term structure (VIX9D / VIX / VIX3M with the curve's own 3-year percentile, DESIGN.md:235); VVIX and SKEW beside it on the same page
- fred_ice_oasRESTATED · never prune · export restricted
- core.fred_observations - the credit axis of Page 10's Regime Monitor. DESIGN.md:145 names 'credit (HY OAS z)' verbatim among the raw axes the regime label is computed from at read time, and DESIGN.md:171 uses price-vs-credit as one of the standing cross-source divergence tests
- fred_tier_aRESTATED · never prune
- core.fred_observations - Page 11's macro browser and its REVISION TRAIL, which DESIGN.md names as the differentiator over v1. Every backtest that keys on a macro release reads these through the point-in-time predicate rather than today's value: PAYEMS has been revised on 1,051 of its 1,052 observation dates, so the difference is not cosmetic
- metric_dailyRESTATED · export restricted
- core.metric_daily - the Data Explorer and the MCP server (DESIGN.md:397), and every context-helper percentile, robust z and recurrence interval behind them (DESIGN.md:143). DESIGN.md:131 calls this the foundation that must land before any page beyond the Data Explorer
- occ_open_interestRESTATED · never prune · export restricted
- core.occ_open_interest - Page 5's participation series: the market-wide equity-versus-index open-interest split and each leg's own 2-year percentile. DESIGN.md:84 pairs OCC with Cboe for 'daily volume/OI by product class' feeding the participation series, and names the shape assertion this table's cross-foot CHECKs implement
- tiingo_pricesIMMUTABLE
- core.price_daily + core.corporate_actions - the point-in-time price spine read through core.prices_asof(); every return, every relative-volume and dollar-liquidity statistic, and the denominator under borrow rates, multi-leg premium and ETF weights
- uw_borrow_ratesIMMUTABLE · never prune
- core.borrow_rates - hard-to-borrow cost and availability per Core name; the squeeze/crowding leg of the short-interest view
- uw_earningsRESTATED · never prune
- core.earnings_events - the Options page's expiry-window panel (DESIGN.md K: which expiries span the next report, with what buffer), the term-structure earnings overlay (DESIGN.md §4 page 4), the Symbol Dossier's earnings module, and DESIGN.md R's implied-versus-realised base rates, which read the expected move as it stood before each report
- uw_economic_calendarSNAPSHOT_ONLY · never prune
- core.economic_calendar - Page 14 Calendar and Page 2 On The Clock: every scheduled release with the forecast as it stood each day before the print, which exists nowhere once the figure is released
- uw_etf_holdingsIMMUTABLE · never prune
- core.etf_holdings - ETF composition and exposure over time; the sector/country decomposition behind the concentration and crowding views
- uw_etf_profileSNAPSHOT_ONLY · never prune
- core.etf_profile - daily fund state for the tracked ETFs. holdings_count is the vendor's own completeness denominator for core.etf_holdings, which caps at 1,000 rows; aum and the four volume fields are the only in-band freshness signal this ETF lane has, since neither /weights nor /info carries a vendor vintage
- uw_etf_weightsSNAPSHOT_ONLY · never prune
- core.etf_weights - the look-through sector and country decomposition behind Page 13's concentration view and the effective-bet count. DESIGN.md:43 names rebalance detection, which is the day-over-day diff of these weights; the sector legs of Page 10's regime axes read it too
- uw_flow_alertsIMMUTABLE · never prune · export restricted
- core.flow_alerts - Page 7 Flow Discovery: market-wide unusual flow on tickers NOT tracked, with triage memory (time-boxed dismiss, flare-resurface when premium doubles past the dismissal snapshot); and Page 6 Flow->OI: each alert's contract joined to the next session's open-interest change, with the consecutive-session persistence counter GOVERNANCE §5 defines. DESIGN.md W grades the alerts themselves as an external signal
- uw_interpolated_ivRESTATED
- core.iv_interpolated - Page 4's Chain & IV Surface, which DESIGN.md:113 names as the page the owner asked for by name, and DESIGN.md:157's IV rank versus true percentile badge with its divergence flag. It is also the fix DESIGN.md:151 calls the single highest-value one in the document: v1 stored one front-month IV that collapsed to the 0DTE reading every day SPY listed one, which is every day
- uw_multilegIMMUTABLE · never prune
- core.multileg_trades - spread and condor reconstruction; the structured half of options positioning that single-leg flow cannot show
- uw_newsIMMUTABLE · never prune
- core.news_headlines - Page 15 News (volume and sentiment by source, novelty scoring by cluster against its own history) and the Symbol Dossier's news module, keyed through the tickers array
- uw_option_chainsRESTATED · export restricted
- core.option_strikes - Page 4 Options Chain & IV Surface: call/put OI, OI change against prev_oi, side-split volume and gamma on one strike axis per expiry, and the GEX and max-pain derivations DESIGN.md L measures
- uw_predictionsSNAPSHOT_ONLY · never prune
- core.prediction_snapshots - DESIGN.md M's options-implied versus prediction-market-implied probability for the same event (Fed decisions, elections): one price per Finance/Politics market per day, from first sighting to resolution
- uw_risk_reversalIMMUTABLE · never prune
- core.risk_reversal - DESIGN.md:159's 25- and 10-delta risk reversal, each as a percentile of its own history, and the skew small-multiples on Page 4's options workbench (DESIGN.md:402)
- uw_variance_risk_premiumIMMUTABLE
- core.variance_risk_premium - DESIGN.md:159 calls it the most decision-relevant volatility statistic available and absent from every survey's volatility suite. Page 5's Vol & Dealer Regime, and the implied-versus-realised leg of the options workbench
- xnys_calendarIMMUTABLE · never prune
- core.trading_sessions - every date-parameterised backfill, the (source,date) coverage view, and every forward-return window