UK region-grain activation: fan out the 18 two-level targets over the region tier and make region a cross-grain control (#905)

changelog.d/905-uk-region-grain-activation.added.md

@@ -0,0 +1 @@
+Activate UK region grain: the 18 two-level (country + region) contract targets — ONS population by age band and VOA council-tax stock by band — fan out over the twelve-area region tier (nine English regions, Wales, Scotland, Northern Ireland), one reference per area scoped on the spine by a household-region predicate, replacing the 18 single country rows; the cross-grain rule gains `region` between country and constituency with every tier code as its own leg, resolved for English areas through the crosswalk's ladder-derived region membership (microcosm#905).

changelog.d/uk-890-target-binding.added.md

@@ -0,0 +1 @@
+Bind the transport and energy facts of chronicle #254/#257 on the UK national surface (#890, the target-binding half of the #790 "bind facts and calibrate" ruling). The DfT London bus rows bind at region level through a region-aware generator pin and a compile-time roster check; England outside London is derived from the bound England and London rows; BUS05ai fare receipts are aligned from the year ending March to the calibration calendar year with the BUS0415 fares index (England ×1.057, London ×1.0; net support never aligned), with the factor and its source record ids declared on the spec metadata. Eight net-new national targets bind pinned facts: OBR fuel duty receipts from cars on the engine's cars-only `fuel_duty` (the all-road-users `obr.fuel_duties` row is signed out and its #757 measure exclusion retires), ORR table 7270 rail support (Great Britain) on `rail_subsidy_spending`, ONS Consumer Trends COICOP 04.5 on `domestic_energy_consumption`, and the Scotland, Wales and Northern Ireland local-bus legs (Scottish Transport Statistics passenger revenue and government support; StatsWales outturn support through local authorities; DfI Translink receipts; NITHC public transport support as a joint bus-plus-rail cell). Contract 236 → 244 targets, active references 415 → 424, national compile 424/0 unsupported. One dated reviewed target-fit exclusion (`obr.fuel_duties_cars@2025`, −25.6% on the spine-p baseline, expires 2026-10-11) defers the release fence until the spine half lands; the baseline receipts in `experiments/890-fuel-bus-energy-receipts.md` record that every other new row fits by weight stretch alone, which is what the spine half is built to remove.

changelog.d/uk-chronicle-repin-890.changed.md

@@ -0,0 +1 @@
+Re-pin the UK national target surface to Chronicle `c6f9361` (#890, from `ec7169b`). The artifact carries the 16 packages of chronicle #254/#255 and the six of #257/#258 (138,847 rows, +7,397): HMRC hydrocarbon oils quantities, OBR fuel duty receipts by vehicle category, DESNZ pump prices, DfT licensed cars by fuel type, Scotland/Wales/NI bus finance, ORR rail finance and fares index, NEED consumption tables, Ofgem cap levels for every charge restriction period and region, ONS Consumer Trends COICOP expenditure with `coicop` as a row dimension on both sheets, and NTS0313/NTS0621 frequency of local bus use. The chronicle consumer-fact schema digest moved (`76ac268e…` → `72ad3149…`) because `quarter` and `week` period types were added; no existing UK national reference resolves a different fact or value (415 active, 7 deferred, 8 signed out, unchanged), so the committed references and membership regenerate byte-identical apart from the feed identity. The local census, local validation levels and local reference surfaces keep their separately reviewed `ec7169b` pin. Adds `tools/vendor_uk_ledger_facts.py` with `uk/ledger_fact_vendor_selections.json`: pinned Chronicle facts the spine stages and target diagnostics consume are now copied verbatim into per-concern resources (`dft_bus_value_anchors.json`, `road_fuel_anchors.json`, `licensed_cars_fuel_type.json`, `need_energy_facts.json`, `ofgem_price_cap_facts.json`, `nts_bus_use_frequency.json`, `devolved_bus_finance.json`, `orr_rail_facts.json`, `ons_household_expenditure_facts.json`), each recording the feed identity it was taken from and regenerating byte-identical from the pinned feed under test; no value in a vendored resource is derived.