Ratebook v0.1.0
First tagged release of Ratebook — the open rate engine for the electrified home.
What's in it
- Deterministic rate engine (Python + a TypeScript port held to it byte-for-byte by shared vectors), cross-validated against NREL's PySAM. Prices flat/tiered/seasonal/TOU residential tariffs; "unknown" is a typed refusal, never a wrong number.
- v0 tariff dataset — 65 engine-validated tariffs across 37 US utilities / 26 states + DC, URDB-derived and CC0, with provenance + confidence per record (
packages/ratebook-data/dataset/). - MCP server — 4 tools (
lookup_tariff,estimate_bill,compare_plans,best_charge_window). - Home Assistant integration — live price + cheapest-charge-window sensors, self-contained (no PyPI/network deps), with a 'pick your utility' bundle of 15 plans + a reconfigure flow. HACS mirror: cbetz/ratebook-homeassistant.
- Accuracy scorecard — published golden-set eval (
docs/GOLDEN_SET.md).
Code Apache-2.0; published data CC0-1.0 (derives from the CC0 U.S. Utility Rate Database).