32  Data Catalogue

32.1 Data provenance

The full domain/dataset reference already lives in platform/domain-datasets-reference.md — 34 datasets across 4 domains (eod, quarter, annual, eow), sourced from FMP, FRED, and FINRA, with Bronze/Silver/Gold table mappings and per-dataset quality flags. That document is generated from the platform’s own config (config/dataset_keymap.yaml) and should stay the single source of truth for provenance rather than being duplicated here.

For the research specifically (beyond what the trading platform ingests), the research proposal’s data section also names sources not yet in the platform: SEC 13F institutional holdings and SEC 10-K supply-chain parsing, needed for the ownership and supply-chain network layers. See the gap analysis and research roadmap for which of those are already built and which are proposed but not yet built.

32.2 Licensing

  • licensing.md — per-vendor terms (FMP, FRED, FINRA, and any research-only sources added for the network layers).

32.3 Refresh schedules

  • refresh-schedules.md — cadence per domain, drawn from the dataset reference’s per-dataset Refresh column.