# ELECTION-01 — Hub product proving ground

Separate research branch for honest public-data aggregation + point-in-time backtesting of US presidential **national two-party popular-vote margin**.

**Not** an SRP scoring change. L1/M1 authorizations untouched.

| Artifact | Path |
|----------|------|
| Protocol draft | `PROTOCOL_V0_1_0_DRAFT.md` |
| Protocol fields JSON | `protocol_fields_v0_1_0.json` |
| Integrity contract | `INTEGRITY_CONTRACT.md` (sha256 `145046ff15faa3f6d71e07c817f1d68b4a8addcd7bafea6dbb98a44a0ba4227b`) |
| Coverage/release ledger | `COVERAGE_RELEASE_LEDGER.md` + `coverage_release_ledger.json` |
| Attempt ledger | `attempt_ledger.json` |

Hub mirrors: `/docs/election01-*` on ai-hub after deploy.

## Integrity negative controls (executable)

See `NEGATIVE_CONTROLS.md` and `negative_controls_report.json`.
Run: `python3 scripts/election01/run_negative_controls.py`
MODEL_CHANGE=NO — stub/frozen predictor only; no fitting claimed.