Warconomy gives analysts a versioned, source-linked dataset with machine-readable exports, provenance, deterministic checksums, and reproducible version pins. Pull the export, pin a version, check provenance and confidence, and compare only like units. Values are market benchmarks tracked alongside risk — not causal attributions — and coverage is partial, not real-time.
- Machine-readable exports + reproducible version pins.
- Provenance, confidence, and freshness on every figure.
- Compare like-for-like; benchmarks, not causal claims.
Pull the machine-readable export
Start from data.json, or the CSV/JSONL distributions for spreadsheet/notebook work. Download bundle.
Pin a version
Freeze a release so a model or report is reproducible; the registry lists every materialized version. Version registry.
Check provenance and coverage
Every figure links to its source; the coverage report shows how much is source-linked. Provenance coverage.
Read confidence, not just value
Confidence and review dates qualify each figure. Treat low-confidence or stale values accordingly. Quality scorecard.
Compare like with like
Rankings compare a single shared metric/unit; do not mix units. Benchmarks are tracked alongside risk, not causal attributions. Comparisons.