[policyengine]

Automated certification candidate for policyengine-uk 2.141.10, policyengine-core 3.33.0 and uk-data 1.59.1 (enhanced_frs_2024_25, sha256 e8483bd8453b4682…).

• Pair id: ff159b50f9e76616af230cdfca421821cee438e162f68e0281274632703085a5
• Ledger: https://github.com/PolicyEngine/policyengine.py/tree/uk-qualification-ledger/uk/ff159b50f9e76616af230cdfca421821cee438e162f68e0281274632703085a5

Gate	Status	Summary
installed_versions	pass	policyengine 6.9.0, policyengine-uk 2.141.10 and policyengine-core 3.33.0 are installed and their requirements hold.
wheel_hashes	pass	The pinned PyPI wheels match their RECORDs and the installed files (policyengine-uk: 2964 wheel members, 2964 installed files, policyengine-core: 199 wheel members, 199 installed files).
core_compatibility	pass	policyengine-core 3.33.0 satisfies policyengine-uk>=3.32.27; certification will record an exact qualified pair, since the publisher's claims do not cover both.
data_access	pass	Read the release manifest and 7 artifacts with the configured credential; anonymous access returned HTTP 401.
dataset_bytes	pass	enhanced_frs_2024_25 is the pinned 113,961,901 bytes with the UK single-year layout.
supported_years	pass	Supports 2024–2032, covering the required and horizon years.
budget_inputs	needs_review	11 model inputs are newly missing from the data (additional_residential_property_purchased, is_full_year_category_a_director, loss_relief, non_residential_property_purchased, non_residential_property_secured_debt, other_residential_property_secured_debt, pension_annual_allowance_taxable_lump_sum_death_benefits, pension_contributions_via_salary_sacrifice_pre_2015…)
reform_contract	pass	Contract 2.1.0: every case, property and population check passed (905); 34 pending a methodology call; 120 unsupported by design.
metrics_review	needs_review	model_only: checks newly warn: inert:war_widows_pension; the model's fiscal partition changed; metrics have no baseline: caseloads.war_widows_pension 2025, caseloads.war_widows_pension 2026, caseloads.war_widows_pension 2027, caseloads.war_widows_pension 2028, caseloads.war_widows_pension 2029, caseloads.war_widows_pension 2030, caseloads.war_widows_pension 2031, spending.war_widows_pension 2025; 153 metric move(s) exceed their review bounds; data_only: checks the previous pair ran were not run: sign:sda, sign:dfe_education_spending; 150 metric move(s) exceed their review bounds; 329 metric change(s) are withheld because a figure from fewer than 10 records is suppressed on one side, or the two sides leave out different ones and cannot be lined up
reform_responsiveness	pass	All 10 standard reforms ran and moved their targets by at least their minimums in every year from 2025-26 to 2031-32.
replay	needs_review	Replay moves need review: data_limited_exception 5, flagged_move 404

Verdict: needs_review. Merging this PR promotes the pair. Until cos d1344 is ruled, promotion is a reviewed merge; the certification evidence test refuses an exact-pair certificate without a promotable ledger receipt.

🤖 Generated with Claude Code

[policyengine]

This pair is not on main's data line, so it is never promoted. The ledger keeps this candidate's evidence.