Feature store catalogue

FEATURE-STORE CATALOGUE — the derived research panels this platform computes for its own models: what exists, how much of it there is, how far back it goes, and how fresh it actually is. Every entry carries MEASURED coverage (rows, symbols, distinct observation dates, history window) and a measured freshness block. freshness.status is current only when the producer is inside twice its own declared cadence AND the data is on its own rhythm, and archival when either fails but the dataset is a genuine historical panel — an archival dataset is still readable, EVERY read of it says so, and not_current_reason names the clock that is behind (data_behind_its_own_cadence for a producer that runs over old data). stale_for_its_own_cadence fires when the newest observation (or a panel's newest vintage, freshness.vintage_clock) is more than three of that clock's own typical gaps old — counted in weekdays for a daily panel, so a weekend or holiday is not a missed observation. vintage_column names the per-row vintage a panel carries (ratio_vintage, returns_data_through); coverage.rows_dated_after_today counts rows stamped with a future date and, on a monthly panel, coverage.rows_written_before_month_end the rows written before their month ended (a month in progress); neither is served. Datasets whose producer stopped and which are NOT panels are listed under withheld with the measurement that disqualified them and have NO route at all; a dataset MEASURED into that state (not a panel, and its producer stopped or its data behind its own rhythm) carries freshness.status withheld and is refused on read; excluded lists live datasets deliberately not sold here, with the reason. Live-measured 2026-08-02: 19 servable datasets totalling 17,062,613 rows — 14 current, 5 archival (spanning 1962, 1970, 2000, 2025-05 and 2026-05 forward) — plus 5 withheld and 2 excluded. Call it first: it is the only place the slugs for /api/v3/features/{dataset} are published.

A 200 is a JSON envelope: ok: true and this endpoint's own fields (see Research Datasets). It carries X-Request-ID, the credit headers (X-Credits-Cost, X-Credits-Remaining, X-Credits-Receipt, X-Credits-Settlement) and the rate-limit headers.

A 200 is billed even when it holds no data (an empty list, null, found: false), so check those fields. Pricing & credits explains the credit headers.

No error is billed. The body is { "ok": false, "error", "detail", "request_id" }; read the code from error.

This table lists what the spec declares for this route. Error responses lists every code FIRM sends, with when to retry.

Read the Research Datasets guide for this endpoint's fields and examples.