Journal branch CI red since 2026-08-24: 9 unreviewed identity mints (incl. raw-id concept bls.cps.LNU02374597) vs the reviewed anchor; blocks #182

The journal branch's own CI (Arch checks) has failed on every push since 2026-08-24 (runs on 54dbabc8, 734beb85, 9526211a, d77afe26). Two tests in tests/test_build_series_catalog.py are red, and both are the test doing its job:

1. test_committed_catalog_is_current_and_valid: committed["suspect_segments"] == ['LNU02374597']. The catalog now carries a series whose concept is spelled with a raw BLS series id, bls.cps.LNU02374597 (uuid 34af6867-ce27-468a-a49b-e9a1dc9dfef2, first observed 2026-07, 1 observation). The builder flags id-shaped segments for curation by design; the fix is a semantic concept spelling (bls.cps.<what it measures>) via the documented --allow-remint --remint-note path, decided by whoever owns the docket admission.
2. test_identity_uuid_map_matches_reviewed_anchor: reviewed anchor c0b8a9c7… (set 2026-08-13 in 1a38336, 219 series) vs current 3ce86d62… (228 series). Since the resolver's appends regenerate the catalog in the same commit (the #174 follow-up), the daily Record N first-print observation(s) commits on 8/23–8/28 minted identities without a reviewed anchor update. Exact delta, computed from the two catalogs:
   • 9 new UUIDs: ssa.ssi.recipients.colorado.aged_65_plus, ssa.ssi.recipients.colorado, ssa.ssi.recipients_aged_65_plus, ssa.oasdi.disabled_worker_beneficiaries, bls.cps.LNU02374597, bls.cps.lfpr_55_plus, bls.laus.colorado.labor_force, bls.ces.home_health_care_services.employment, va.vba.mmwr.claims_inventory (all country / economy-aggregate).
   • 7 placeholder enrichments, UUIDs preserved (geography/entity null → country 0100000US / economy aggregate): fed.g17.capacity_utilization.manufacturing, fed.g17.manufacturing_production_mom, census.housing.permits_saar, census.housing.completions_saar, census.new_residential_sales.new_single_family_houses_sold_saar, bls.export_prices.all_commodities_mom, ssa.ssi.total_recipients.
   • No removals, no supersedes; render() matches the committed bytes (the catalog is coherent, the review pin is what is stale).

What this blocks

Arch checks is a required status on codex/thesis-ledger-facts (strict), so every PR into the branch is unmergeable until it is green — currently #182 (OTS anchoring; the proofs already carry Bitcoin attestations at blocks 963253/963257), whose head is otherwise clean.

Proposed resolution (Thesis lane, since these are docket identities)

1. Review the 9 mints; rename bls.cps.LNU02374597 to its semantic concept with --allow-remint --remint-note, or retire it if the admission was unintended.
2. Update the anchor constant and len(committed["series"]) in the test with the dated justification comment, in the same diff, per the test's own rule.
3. Process fix on the Thesis side so automatic appends cannot mint identities without an anchor update landing in the same commit (or the mint path opens a PR instead of pushing).

Source of the appends: ThesisInstitute/thesis scripts/resolve_pending.py via its scheduled workflows.

Start with tests/test_build_series_catalog.py and run the Arch checks covering the two failing tests. Read scripts/resolve_pending.py and compare the reviewed and current catalogs to review the nine mints and the stale anchor. Done means the identity decisions, catalog anchor, and series count are updated with the required justification and Arch checks pass.