Build N epic: CD target surface + base rebuild bundle (quartet, assets, QRF fix, feed regen, gate hardening)

[MaxGhenis]

Max greenlit (2026-07-17) the congressional-district target surface as the next build after the Build M certification: "yes do it."

What

Turn on --include-congressional-district-targets for the next certified US build (Build N), expanding the calibration surface from the national+state class (~5,667 compiled targets in Build M) to the full geographic feed surface: the v8 Ledger feed carries 23,968 congressional-district facts alongside 11,191 state and 2,101 national — a ~30k-target-class surface.

Mechanics already in place

• The builder flag exists and is opt-in-off; the compile path translates CD facts via --congressional-district-vintage-crosswalk (source vintage → current districts) and hard-requires household.congressional_district_geoid on the support frame — which the Build M frame already carries (export parity tracks it at +2.3% drift).
• gate_congressional_district_targets currently records False in build manifests; Build N flips it to a gated surface.

Work items

1. Crosswalk provenance: pin the source-to-119th crosswalk artifact (sha-pinned input, Build the 117th->119th CD vintage crosswalk from Census sources #288 alignment); the checkpoint identity and materialization-cache context already carry its sha.
2. Preflight sizing (no solve): compile the CD-included registry, run the Add release-gate preflight: catch statically-knowable gate failures without a solve #436 preflight zero-support/support-expressibility pass over the 57k sparse selection — CD cells are exactly where a frozen sparse support goes thin (the Build G Build E: certification-candidate staging build from post-#287 main (fix target_compilation OOM first) #299 exclusion class). Expect a batch of per-cell reviewed exclusions or a reselection decision; adjudicate BEFORE burning a run (the campaign's attempt-9/10/11 lesson, now tooled).
3. Solve-cost calibration: ~5.4× more constraint rows; measure a bounded-epoch dense probe before committing the full ladder. Loss-weighting review for the CD family (11k state rows already dominate counts; 24k CD rows will need family-level weighting the same way state rows got it).
4. Gate adjudication pass: within-10% and zero-support baselines will move on a 30k surface; set expectations from the probe, not Build M's 90.2%.
5. Scorecard: report Build N vs Build M on the shared national+state subset AND the CD surface separately, so the flip decision compares like with like.

Sequencing

After the Build M cert (policyengine.py#475) lands. Interacts with the geography-ladder/ACS lane (#288, Build L plan): the CD vintage story should be settled once and shared between calibration targets and the geography assignment.

[MaxGhenis]

Max (2026-07-17): time the CD surface's cost, and "if not too much we'll remove the flag and make it default behavior." So #449's deliverable includes a decision-grade timing comparison, and the flag's fate is explicit:

Timing protocol (decision evidence)

Run the same commit, same base, same machine, no other compute, and record wall-clock per stage from the builder's own stage telemetry plus /usr/bin/time -l peak RSS:

stage	CD-off (Build M baseline)	CD-on
registry compile	~2 min (measured tonight)	?
target-frame materialization (cold)	?	? (per-target cache shared where applicable)
constraint-matrix build	?	?
sparse refit (1500 epochs, frozen selection)	?	?
dense solve	?	?
gate battery + export + smoke	?	? (should be ~unchanged)
total sparse arm	~4.5 h under contention / ~3 h clean	?

Also record: peak RSS per stage (jetsam headroom on 128 GB is a live constraint), and loss/within-10% on the shared national+state subset so quality regression is visible next to the cost.

Decision rule (Max's)

If the CD-on delta is acceptable ("not too much"), remove --include-congressional-district-targets entirely and make CD targets default behavior — no permanent flag, no config fork; the surface becomes the surface (consistent with the no-codified-tiers stance). The vintage crosswalk then becomes a required pinned input of every US build rather than an optional pairing.

If the delta is material, keep the flag and record the measured cost here for the next revisit.

[MaxGhenis]

Scope ratified by Max (2026-07-17). Build N's theme: one base rebuild + one feed regeneration + the CD surface + gate hardening. Build O's theme: the ACS multispine merge.

In (Build N)

workstream	issues	why bundled
CD target surface + timed A/B (this issue)	#449, fold #314	crosswalk becomes a pinned required input; flag deleted if cost acceptable (Max's rule)
Groundwork readiness quartet	#451, fold #340/#361 residuals	the two structural zeros force the base rebuild; tips/overtime need feed facts + targets + protections
Asset magnitudes	#374	SIPP blend rides the same base rebuild; SSI-aware refit rides the solve; oldest slipped promise on the list
Partnership-SE QRF fix	#432	its parity exclusion (#433) explicitly lifts with the rebuild — this is the rebuild
Feed regeneration fixes	#410, #411	load-bearing the moment the new facts (tips/overtime/SE-health/SNAP) regenerate the feed; #411 is a tips-family fence bug
SNAP caseload targets	#370	same target-sourcing motion as the quartet
Coverage effective-mass semantics	#396	the keogh weight-crush lesson: zero-weight signal must not count as restored
Dataset-role classes	#398	enabler for first-class non-default publishing (M dense, ACS local)
Release-delta consumer report	#366	the client-facing artifact for every future flip (Groundwork asked for exactly this in effect)

Conditional: #424 (SSI candidate undercount) jumps in if the dense #447 trajectory adjudication points at the candidate universe.

Out (Build O)

Multispine program (#390, #395, #403, #346) and modeling-layer items: #372, #341, #321, #327, #417, #320, #360, #416, #412.

Parallel tracks (not build-gated)

Architecture/executor: #378–#382, #326, #388. UK/BE/NZ lanes run independently.

Hygiene

Verify-and-close sweep of superseded issues (#377, #386, #359, #356, #434; #367 stays open — unexercised, not fixed) running now with per-issue evidence.

[MaxGhenis]

Scope addition from the SSI asset-limit analysis loop (cross-checked on the certified Build M):

• Child SSI structurally absent: ages 0–14 have no disability signal (baseline child SSI 0.12M vs SSA 0.98M); seed qualifying disability per the #430 pattern #453 (child SSI) joins the Build N bundle, riding the same SIPP donor stage as SIPP+SCF blend + SSI-aware refit: land the asset magnitudes for the Build J re-certification #374 — one stage closes both. The under-18 baseline gap (0.12M vs SSA 0.98M) is the entire remaining SSI baseline miss (18–64 and 65+ are within ~2%), caused by is_disabled ≡ 0 for ages 0–14 (CPS battery starts at 15); SIPP pu2023 carries verified-populated parent-reported child disability items (ECOGNIT 100% answered / 9.2% yes for 5–14, plus hearing/seeing/ambulatory/self-care) and RSSI_YRYN as a take-up anchor, with 0–4 banded to the SSA age curve.
• SIPP+SCF blend + SSI-aware refit: land the asset magnitudes for the Build J re-certification #374 acceptance sharpened by the certified-Build-M measurement: $10k/$20k currently +$5.73B/+793k (unchanged from Build J — the asset stage is still SCF-only: 64.3% zeros, mean $119k); dense-class acceptance ~+$1.60B (OACT $1.0B / CBPP 224k). The $2k–$20k band mass is the lever; full-removal is already robust.
• Acceptance harness available as a build gate: ~/PolicyEngine/ssi-asset-limits/.local-results/populace-rerun-scripts/ (pass_c_scenario.py + pass_f_combine.py, ~5 min/scenario on any 57k artifact; pass_h_* = the 40k-chunked variant for dense frames). Anchors: baseline SSI by age vs SSA 3.99M/2.42M/0.98M; $10k/$20k vs OACT $1.0B / CBPP 224k (23k children / 154k adults / ~47k elderly).
• Build O note (ACS-multispine staging: SSI probe finds spine-inconsistent incidence (2.6% vs 1.9% baseline rate) and weight-composition skew (ASEC 1.5 persons/hh, ESS 0.3%) #403): cheapest first multispine experiment per the runbook's decision-5 — re-solve with l2>0 + heavier population-target weights against the spine-inconsistent SSI incidence (ACS 2.59% vs ASEC 1.85%) and the CA/NY/NJ/PA population misses.

[MaxGhenis]

Scope addition from the IARIW OBBBA paper loop: #454 (ESI employer premium contributions) joins the Build N bundle, same workstream as the #451 quartet and the #340/#361 residuals — it is that exact class at ~100× the mass: dense f0af251 imputes ~$1.13T (right scale vs KFF), sparse exports drop the column, nothing anchors it, and the frozen eCPS computed $0 (documented as a levels-omission in the IARIW paper's CBO-style resource measure). Fix ladder is the standard motion: input-coverage requirement across release classes, Ledger fact for aggregate employer contributions (NHEA employer share or KFF EHBS × covered lives), calibration target + release gate against the anchor. The per-column disposition pass over the #361 94-column inventory should treat ESI as its highest-mass member.

[MaxGhenis]

CD crosswalk revival done: PR #288 rebased onto main tip (86871cf, post-#455), conflicts resolved (builder import list; CD test now asserts the packaged-default semantics), 14 crosswalk + 116 builder tests green locally, CI green on 3.13/3.14. Marked ready for review — this unblocks the CD target surface (#314) and the timed CD A/B.

[MaxGhenis]

Scope addition: #463 (stage the release builder — the #406 pattern applied to build_us_fiscal_refresh_release.py) joins as the dense-lane front item, replacing the closed #456. Tonight's full-scale dense measurements are in the issue: #461 fixed the leak class (RSS returns between families), and the remaining legitimate working set crests at ~429GB / ~536GB (eligibility high-water, SSI reconcile passes) — past physical RAM it swap-crawls (6h of zero log writes at ~1 effective core), which is a throughput wall, not a correctness one. Acceptance: dense full-pool on n2-highmem-32 with the #406-style bit-equivalence proof. The #406 checkpoint machinery is directly reusable; prior art codex/jct-materialization-memory-20260617.