Infrastructure intelligence sample set

Schema and provenance

Generated synthetically (seeded, reproducible) for the framework prototype. Facility rows carry coordinates, sector, operator, capacity, and status; advisory rows carry a facility reference, kind, severity, and dates. No record describes a real facility, operator, or advisory. The season time series accumulates as the engine runs; its checksum is intentionally omitted because the file grows daily by design. The immutable artifacts are the per-day snapshots and their metadata.json manifest.

Facility fields: id, name, sector, operator, lat, lon, capacity, unit, status. Advisory fields: id, facility_id, kind (outage | maintenance | alert), severity (low | moderate | high), summary, issued, resolved. The time series (timeseries.csv) carries one row per snapshot date: facilities, sectors, active_advisories, facilities_affected.

Full daily snapshots and the archive metadata.json (with per-snapshot SHA-256 checksums) live under the engine’s snapshots/ directory and mirror into the dashboard bundle. The time series is excluded from verify:datasets because it is append-by-design, not immutable.