Skip to content

Point the IN-MT source-conflict candidates at their code on main - #10351

Merged
MaxGhenis merged 6 commits into
mainfrom
source-conflict-candidates-in-mt
Oct 25, 2026
Merged

MaxGhenis merged 6 commits into
mainfrom
source-conflict-candidates-in-mt

Conversation

@MaxGhenis

@MaxGhenis MaxGhenis commented Oct 21, 2026 •

Copy link
Copy Markdown
Contributor

Brings the Indiana to Montana entries in source_conflicts/candidates.yaml up to date with the code on main, and adds candidates for conflicts the merged code's own comments record. One of three PRs splitting this refresh by state; the follow-up promised in #10349.

Why. #10287 wrote these candidates while their implementing PRs were still open, so each said code: [] or "Code not yet on main". A PR each one names has since merged: #7887, #8902, #8942, #8982, #9180, #9213, #9214, #9218, #9240, #9250, #9267, #9297, #9379, #9380, #9671, #9746, #9791, #9796, #9840, #9900, #9903, #9915, #9945, #10024, #10279, #10281.

Since the first review. This head addresses the GPT-6.1 Sol review and @DTrim99's review:

follows flips. Each changed because the merged code does something different from what the candidate recorded:

Refreshed candidates (29). Each now lists the tracked files that decide it under pe_us.code. Its note says which source that code follows, as read on main, and quotes nothing the code does not show. pe_us.follows stays unknown wherever the readings begin "No statute supplied".

Candidate Code What the code follows
us-in:eitc-current-versus-frozen-federal-base in_eitc.py, in_eitc_eligible.py PE-US follows Schedule IN-EIC: under static conformity in_eitc multiplies the current-year federal credit by the match rate, and its comment argues the post-2023 federal changes are inflation indexing only (conflicting-source)
us-in:rcap-unlicensed-facility-per-diem unlicensed.yaml, in_ssp_rcap.py PE-US uses $37 from July 1, 2008 and $27 before, citing 455 IAC 1-3-3 and the September 2017 FSSA RCAP Provider Guide (unknown)
us-ks:medicaid-parent-dependent-child-age age_limit.yaml, student_option.yaml, is_parent_for_medicaid_nfc.py PE-US follows the manual: Kansas's age limit is 19 from 2014 and its student option is off, and the age-limit comment records the S25 student-option election (unknown)
us-ks:snap-crop-share-farm-rent-classification snap_farm_rent_income.py, unearned.yaml PE-US follows the federal regulation in every state, Kansas included: Form 4835 farm rent of a landowner who does not materially participate counts as unearned income, with losses floored at zero (unknown)
us-ky:ktap-payment-maximum-2025 payment_maximum.yaml, ky_ktap_payment_maximum.py PE-US follows the manual from November 1, 2025 ($242 to $627 for sizes 1 through 7 or more), and the parameter's comment records that 921 KAR 2:016 still shows the 2023 schedule (unknown)
us-la:cpi-u-indexing-method amount.yaml, cap.yaml, uprating_extensions.py PE-US follows the bulletin's gap-fill: 2026 uses RIB 26-019's amounts, and later years apply BLS's December-over-December CPI-U change rounded to one decimal and double the rounded single deduction for joint, head of household and surviving spouse filers (gap-fill)
us-ma:circuit-breaker-snap-income disallowed_deductions.yaml, ma_scb_total_income.py PE-US follows the statute's item 2: SNAP and WIC are left out of the add-backs, and the list's comment records DOR's contrary guidance (statute)
us-ma:connectorcare-income-bands-2026 fpl_floor.yaml, fpl_limit.yaml, ma_connector_care_eligible.py PE-US follows the Overview for 2026: income must be from 100% to 400% FPL, and the floor's and limit's comments record the regulation's Plan Type 1 and 3D bands, which are left unmodelled (unknown)
us-md:meap-garrett-oil-level-four md_meap.py, garrett_multiplier.yaml, oil.yaml, regional_counties.yaml PE-US follows the plan's multiplier: Garrett County, the one regional county, pays 1.1 times the statewide amount, rounded half up, so oil Level 4 is 1.1 times $980, or $1,078, and the documentation records the FY27 table's $1,210 in that one cell (unknown)
us-md:meap-interest-dividends unearned.yaml, md_meap_countable_income.py PE-US follows COMAR and the manual: interest and dividends are countable unearned income, and the documentation records that the plan leaves them unchecked (unknown)
us-md:meap-lump-sums-americorps md_meap_countable_income.py, unearned.yaml PE-US follows COMAR: nonrecurring lump sums and AmeriCorps/VISTA payments are not countable, and the documentation says the manual and plan count them (unknown)
us-md:meap-nonqualified-child-household-size md_meap_household_size.py PE-US follows the manual: children under 18 count in household size whatever their immigration status, while ineligible adults do not (unknown)
us-md:meap-strike-pay unearned.yaml, md_meap_countable_income.py PE-US follows COMAR: strike_benefits count, assumed not employee-funded, and the documentation records that manual Attachment B excludes funds without employee contributions (unknown)
us-md:premium-assistance-200-fpl-endpoint final.yaml, initial.yaml, md_premium_assistance_target_contribution_percentage.py PE-US follows the Board design and the MIA exhibit: the 150-200% band ends, and the 200-250% band starts, at 2.00%, and the comment records that the briefing prints 1.00% at that boundary twice (unknown)
us-me:property-tax-credit-unknown-utilities me_property_tax_fairness_credit_utilities_in_rent_amount.py, me_property_tax_fairness_credit_countable_rent.py, utilities_share_of_rent.yaml PE-US fills the gap with the form's estimate: when rent includes utilities of unknown cost it subtracts 15% of rent, from 2014, and the rate's comment says the statute has no unknown-amount estimate (gap-fill)
us-me:sales-tax-fairness-credit-mixed-joint-return me_sales_tax_fairness_credit_eligible.py PE-US follows the schedule's reading: a joint return qualifies unless every filer can be claimed, and the comment says neither 5213-A(6)(C) nor subsection 2 settles the mixed case (gap-fill)
us-me:tanf-incapacitated-adult-care-deduction me_tanf_child_care_deduction.py PE-US follows the rule: the deduction also covers each incapacitated adult needing care, at the $175 tier (conflicting-source)
us-mi:household-resources-dependent-amounts mi_household_resources.py, household_resources_all_members.yaml Partly changed by #9796: PE-US now counts only the claimant's and spouse's own income, as the statute does, except the sources on MI-1040CR lines 21, 22 and 27 (Social Security, SSI, railroad benefits, child support, TANF and general assistance), which it counts for every member as amounts received for others, as the form does (conflicting-source)
us-mi:household-resources-premium-double-deduction mi_household_resources.py PE-US follows the form: line 31 subtracts the family's health insurance premiums less the self-employed health insurance deduction already taken on line 30, so each premium dollar is subtracted once, and Social Security and railroad benefits stay gross on line 21 with their Medicare premiums on line 31 (gap-fill)
us-mo:medicaid-published-dollar-limit-rounding medicaid_income_level.py, medicaid_optional_senior_or_disabled_income_limit.py, mo_mhf_parent_income_limit.py PE-US fills the gap with the manual's convention: medicaid_income_eligible rounds each monthly MAGI Medicaid and CHIP limit up to the next whole dollar and admits income at or below it, in every state as an approximation (gap-fill)
us-mo:property-tax-credit-rounding mo_property_tax_credit.py PE-US follows the DOR charts: the formula reproduces the published charts, which its comment says round the credit down and the phase-out up, although the statute says the nearest whole dollar (conflicting-source)
us-mo:snap-student-child-care-claim-cap per_household.yaml, meets_snap_parent_exception.py PE-US applies Missouri's one-adult-per-household cap from 2025, a start date the parameter's comment calls a modeling choice, and keeps the uncapped federal reading in other states (gap-fill)
us-mo:wftc-investment-income-boundary mo_wftc_eligible.py, investment_income_limit.yaml PE-US follows the statute in every year: investment income equal to the limit keeps eligibility, and the comment records the 2023 form's "equal to or greater than" wording and says the 2024 and 2025 forms use "greater than" (statute)
us-ms:liheap-110-percent-poverty-floor ms_liheap_eligible.py, fpg_floor.yaml, fpg_year_lag.yaml Changed by #10279: PE-US now follows the statute. A household with income below 110% of the prior year's poverty guideline is income-eligible even above the manual's 60% SMI limit, which the comment finds below that floor from size 11 in FY2026 and FY2027 (statute)
us-mt:elderly-credit-homestead-costs mt_elderly_homeowner_or_renter_credit_property_tax_billed.py, mt_elderly_homeowner_or_renter_credit_gross_rent.py, mt_elderly_homeowner_or_renter_credit_selected_claimant.py PE-US follows the form: the claimant counts the property tax billed to their own tax unit and the rent it paid; only when no one in the tax unit is billed does the household's property tax count, as rent. The selected claimant's comment records the statute's whole-homestead reading, which would give the one claimant the household's whole property tax and rent, as not modeled (conflicting-source)
us-mt:elderly-credit-nonloss-agi-adjustments mt_elderly_homeowner_or_renter_credit_gross_household_income.py, gross_income_sources.yaml PE-US follows the statute: gross household income starts from each member's federal AGI, which keeps the non-loss adjustments, and adds back only the business and capital losses (statute)
us-mt:elderly-credit-refundable-credit-timing mt_elderly_homeowner_or_renter_credit_federal_refundable_credits.py, mt_elderly_homeowner_or_renter_credit_received.py, gross_income_sources.yaml, mt_refundable_credits_before_renter_credit.py, refundable.yaml PE-US uses the current return's federal and other Montana refundable credits (the Montana EITC from 2019) as a proxy for credits received, while the elderly credit itself enters as an input for the amount received during the year (unknown)
us-mt:elderly-credit-rent-free-occupants mt_elderly_homeowner_or_renter_credit_gross_rent.py, mt_elderly_homeowner_or_renter_credit_property_tax_billed.py PE-US follows the regulation and instructions: when no one in the claimant's tax unit is billed property tax, the household's property tax counts as rent, taking the larger of that and rent paid, so a rent-free occupant can qualify (conflicting-source)
us-mt:medical-floor-agi-base mt_medical_expense_deduction_joint.py, state_specific_deduction_applies.yaml, mt_standard_deduction_joint.py PE-US fills the gap with the form: through 2023 the joint medical floor applies to the return's pooled Montana AGI (Itemized Deductions Schedule line 1b), and from 2024 the federal floor on federal AGI applies; the joint standard deduction also uses pooled Montana AGI (gap-fill)

New candidates (5). The merged code records each conflict in its comments; none had a candidate or a register entry.

  • us-in:liheap-income-floor-vs-smi-limit (statute). Can Indiana EAP's 60% SMI limit exclude a household of 15 or more whose income is below the federal 110% poverty-guideline floor? PE-US applies the floor to the prior year's guideline, rounded to cents, with the statute's strict "less than" (in_eap_eligible.py).
  • us-ky:ktap-payment-scale-household-size (unknown). Is Kentucky's K-TAP maximum payment scale indexed by the number of eligible persons, as the operations manual says, or by the number of children, as the KY FACES web app labels it? PE-US follows the manual and regulation: the scale is keyed to assistance unit size, proxied by SPM unit size and capped at seven, and the parameter's comment records the web app's Number of Children label (benefit/payment_maximum.yaml, ky_ktap_payment_maximum.py). The KY FACES page is written for kinship caregivers, whose child-only cases count only the children, so the label may describe that audience; confirm which count the agency applies.
  • us-ma:liheap-income-floor-vs-smi-limit (statute). Can Massachusetts HEAP's 60% SMI limit exclude a household of 30 to 34 whose income is below the federal 110% poverty-guideline floor? PE-US applies the floor to the guideline ma_liheap_fpg already lags, rounded to cents, with the strict "less than" (ma_liheap_income_eligible.py).
  • us-md:meap-income-floor-vs-smi-limit (statute). Can Maryland MEAP's 60% SMI limit for households of 11 or more exclude a household of 29 or more whose income is below the federal 110% poverty-guideline floor? PE-US applies the floor to md_meap_fpg's prior-year guideline, rounded to cents, with the strict "less than" (md_meap_eligible.py).
  • us-md:meap-income-limit-plan-percentages (unknown). Do Maryland MEAP's FY2026 income limits for sizes 1-10 follow the OHEP guidelines' 200% FPG amounts or the state plan's lower SMI percentages? PE-US follows 200% of the poverty guideline for sizes 1-10 and 60% of SMI from size 11. For sizes 9 and 10 it uses 200% of the guideline ($119,300 and $130,300), not the guidelines' $11,004 add-on, which gives $4 and $8 more (md_meap_income_limit.py, eligibility/fpg_rate.yaml, eligibility/smi_min_size.yaml).

follows left unknown beyond the "No statute supplied" convention. These readings name no statute or leave the statutory question open, so statute, conflicting-source and gap-fill would each assert a legal reading the code does not make: us-ks:medicaid-parent-dependent-child-age (no source is a statute); us-ks:snap-crop-share-farm-rent-classification (no source is a statute); us-ky:ktap-payment-maximum-2025 (no source is a statute); us-ma:connectorcare-income-bands-2026 (no source is a statute); us-md:meap-garrett-oil-level-four (no source is a statute); us-md:meap-interest-dividends (no source is a statute); us-md:meap-lump-sums-americorps (no source is a statute); us-md:meap-nonqualified-child-household-size (no source is a statute); us-md:meap-strike-pay (no source is a statute); us-md:premium-assistance-200-fpl-endpoint (the statute is silent, but PE-US fills it with the Board design and the MIA exhibit, which are not a form, so gap-fill as the header defines it does not fit); us-mt:elderly-credit-refundable-credit-timing (no source is a statute).

Verification.

  • policyengine_us/tests/code_health/test_source_conflicts.py: 103 passed at this head. I ran a copy outside the package, so pytest doesn't import policyengine_us.
  • The file is main's plus this PR's entries: every other entry is byte-identical. Each changed entry was re-dumped with the PyYAML settings the file was written with (width 110).
  • Ids stay sorted and unique, and every code path is a file git tracks.
  • New floor candidates: each figure was recomputed from the code's formula and parameters on main, and each plan page was read.
  • The three PRs touch disjoint entries; merged together they apply cleanly. After all three, no candidate says "Code not yet on main" or has code: [] while naming only merged PRs.

Not in this PR. The nine Florida and South Dakota LIHEAP candidates landed in #10349. Candidates whose pull requests are still open keep code: []. The other two parts of this refresh are #10350 (merged) and #10352.

axiom: n/a: source-conflict backlog metadata only; no model change

🤖 Generated with Claude Code

Candidates written while their pull requests were open said "Code not yet
on main". Each now lists the tracked files that decide it and a note on
which source that code follows. Adds candidates for conflicts the merged
code's comments record.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@DTrim99

DTrim99 commented Oct 22, 2026

Copy link
Copy Markdown
Collaborator

PR Review: Point the IN-MT source-conflict candidates at their code on main

PR #10351 (author: MaxGhenis), head df1b6ece99. This is a first review. The PR changes two files: source_conflicts/candidates.yaml (28 Indiana-to-Montana entries refreshed, 2 added) and changelog.d/source-conflict-candidates-in-mt.changed.md. Line numbers refer to df1b6ece99.

Summary

#10287 wrote these candidates while their implementing PRs were still open, so most said code: [] or "Code not yet on main". All 24 PRs the body names have since merged. This PR points 28 entries at the merged code and adds us-ky:ktap-payment-scale-household-size and us-md:meap-income-limit-plan-percentages.

The refresh is accurate:

Two should items:

  1. Missing floor candidates. main now records three more 110%-floor conflicts that have no candidate: Massachusetts LIHEAP (from Apply the federal 110% poverty floor to state LIHEAP programs (DC, MA, MS, NJ, OR) #10279, the same PR as the refreshed Mississippi entry), and Indiana EAP and Maryland MEAP (from Apply the federal 110% poverty floor to Indiana EAP, Maryland MEAP and Riverside County LIHEAP #10281). This is the same class as the FL candidate we asked for on Point the FL and SD LIHEAP conflict candidates at their code on main #10349.
  2. The new Maryland MEAP income-limit candidate attributes the additional-member line to the wrong source. The plan has no such line; the $11,004 add-on is on the OHEP guidelines page. The entry also says the guidelines print sizes 9 and 10, which they don't. The question its note leaves open has an answer: the plan's 46.00% for seven people is $1,593 below the guidelines.

Seven suggestions: stale instructions and null URLs that the code on main already answers; four wrong page anchors; a missing deciding file for the Montana homestead entry; the KY K-TAP scale candidate's sources; the Minnesota guardian-fee entry; one PR-body reason; and one older Maryland conflict.

Status:

Critical (must fix)

None.

Should address

1. Add candidates for the three 110%-floor conflicts that main records without one: Massachusetts LIHEAP, Indiana EAP and Maryland MEAP.

Where: the Indiana, Massachusetts and Maryland blocks of source_conflicts/candidates.yaml. Of the IN-MT states, only Mississippi has a floor candidate (us-ms:liheap-110-percent-poverty-floor, line 3542).

The code comments on main:

The state sources set 60% of SMI with no floor:

Why it is in scope:

Effect. The model is not affected. The backlog would miss three statute-against-guidance conflicts in which PE-US follows the statute.

Fix. Merge main first: Indiana's fpg_year_lag.yaml is not on this head. Then add the three entries below, re-dumped at width 110. Every figure comes from the merged code and its tests (ma_liheap_income_eligible.yaml Case 4, in_eap_eligible.yaml Case 6, md_meap_eligible.yaml Case 8). I checked each floor by hand (hand checks 9-11). The ids sort into place:

  • us-in:liheap-... after us-in:eitc-...;
  • us-ma:liheap-... after us-ma:connectorcare-...;
  • us-md:meap-income-floor-... before us-md:meap-income-limit-....
- id: us-in:liheap-income-floor-vs-smi-limit
  program: liheap
  years: {from: 2025, to: 2027, type: program, note: program year as the program defines it}
  question: Can Indiana EAP's 60% SMI limit exclude a household of 15 or more whose income is below the federal
    110% poverty-guideline floor?
  sources:
  - kind: statute
    citation: 42 U.S.C. 8624(b)(2)(B)
    url: https://www.law.cornell.edu/uscode/text/42/8624#b_2_B
  - kind: manual
    citation: Indiana PY2026 EAP Intake and Operations Manual, Sections 3.1 and 8.3, p.23
    url: https://www.in.gov/ihcda/files/Indiana-LIHEAP-Intake-and-Operations-Program-Manual-PY2026.pdf#page=23
  readings:
    statute: Income below 110% of the poverty guideline cannot by itself exclude the household.
    other: Income must be at or below 60% of SMI, which falls below that floor from size 15 in PY2025 and size 16
      in PY2026 and PY2027.
  example: In PY2026, sixteen people with $106,000 exceed the $104,543 limit but are below the $107,965 floor.
  pe_us:
    follows: statute
    code:
    - policyengine_us/variables/gov/states/in/ihcda/eap/in_eap_eligible.py
    - policyengine_us/parameters/gov/hhs/liheap/fpg_floor.yaml
    - policyengine_us/parameters/gov/states/in/ihcda/eap/eligibility/fpg_year_lag.yaml
    prs: [https://github.com/PolicyEngine/policyengine-us/pull/10281]
  state: unreviewed
  note: PE-US applies the floor to the prior year's guideline, rounded to cents, with the statute's strict "less
    than" (in_eap_eligible.py).
- id: us-ma:liheap-income-floor-vs-smi-limit
  program: liheap
  years: {from: 2022, to: 2027, type: program, note: program year as the program defines it}
  question: Can Massachusetts HEAP's 60% SMI limit exclude a household of 30 to 34 whose income is below the federal
    110% poverty-guideline floor?
  sources:
  - kind: statute
    citation: 42 U.S.C. 8624(b)(2)(B)
    url: https://www.law.cornell.edu/uscode/text/42/8624#b_2_B
  - kind: plan
    citation: FY2026 LIHEAP State Plan item 2.1, p.9
    url: https://liheapch.acf.gov/docs/2026/state-plans/MA_Plan_2026.pdf#page=9
  readings:
    statute: Income below 110% of the poverty guideline cannot by itself exclude the household.
    other: The plan sets 60% of SMI for all household sizes, which falls below that floor from size 30 to 34.
  example: In FY2026, thirty-four people with $216,000 exceed the $215,077.68 limit but are below the $216,865 floor.
  pe_us:
    follows: statute
    code:
    - policyengine_us/variables/gov/states/ma/doer/liheap/eligibility/ma_liheap_income_eligible.py
    - policyengine_us/parameters/gov/hhs/liheap/fpg_floor.yaml
    - policyengine_us/variables/gov/states/ma/doer/liheap/ma_liheap_fpg.py
    prs: [https://github.com/PolicyEngine/policyengine-us/pull/10279]
  state: unreviewed
  note: PE-US applies the floor to the guideline ma_liheap_fpg already lags, rounded to cents, with the strict
    "less than" (ma_liheap_income_eligible.py).
- id: us-md:meap-income-floor-vs-smi-limit
  program: liheap
  years: {from: 2025, to: 2027, type: program, note: program year as the program defines it}
  question: Can Maryland MEAP's 60% SMI limit for households of 11 or more exclude a household of 29 or more whose
    income is below the federal 110% poverty-guideline floor?
  sources:
  - kind: statute
    citation: 42 U.S.C. 8624(b)(2)(B)
    url: https://www.law.cornell.edu/uscode/text/42/8624#b_2_B
  - kind: guidance
    citation: OHEP Income Guidelines FY2026 (updated July 9, 2025), 60% SMI footnote
    url: https://dhs.maryland.gov/documents/OHEP/Advisory%20Board/Income-Guidelines-FY2026-Updated-7.9.2025.pdf
  readings:
    statute: Income below 110% of the poverty guideline cannot by itself exclude the household.
    other: Households of 11 or more are limited to 60% of SMI, which falls below that floor from size 29.
  example: In FY26, twenty-nine people with $186,000 exceed the $185,864 limit but are below the $186,615 floor.
  pe_us:
    follows: statute
    code:
    - policyengine_us/variables/gov/states/md/dhs/meap/eligibility/md_meap_eligible.py
    - policyengine_us/parameters/gov/hhs/liheap/fpg_floor.yaml
    - policyengine_us/variables/gov/states/md/dhs/meap/income/md_meap_fpg.py
    prs: [https://github.com/PolicyEngine/policyengine-us/pull/10281]
  state: unreviewed
  note: PE-US applies the floor to md_meap_fpg's prior-year guideline, rounded to cents, with the strict "less than"
    (md_meap_eligible.py).

Then update the body's "New candidates (2)" and the changelog fragment to match.

2. The new us-md:meap-income-limit-plan-percentages attributes the additional-member line to the plan, says the guidelines print sizes 9 and 10, and leaves open a question the sources answer.

Where: source_conflicts/candidates.yaml:2686-2716.

What the entry says:

  • "statute: No statute supplied; the published guidelines give annual amounts at 200% FPG for sizes 1-10."
  • "other: The plan states truncated equivalent SMI percentages and an additional-member line inconsistent with them."
  • The note: "PE-US follows the published 200% FPG amounts for sizes 1-10 ... Verify the plan's figures and whether they change any household's eligibility."

What the sources say:

  • The plan has no additional-member line. FY2026 plan item 2.1, page 9 lists one SMI percentage per size: 39.00%, 40.00%, 41.00%, 41.70%, 42.00%, 42.40%, 46.00%, 50.90%, 54.89% and 58.70% for sizes 1-10, then 60.00% for sizes 11-15.
  • The add-on is on the guidelines page. The OHEP Income Guidelines FY2026 print annual rows only for sizes 1-8 ($31,300 to $108,300), then "For each additional HH member add $11,004".
  • That add-on differs from 200% FPG. 200% of the $5,500 increment is $11,000. So sizes 9 and 10 come to $119,304 and $130,308 on the page, against $119,300 and $130,300 at 200% FPG.
  • The code follows 200% FPG there. The merged test says so (md_meap_income_limit.yaml Case 1): "Sizes 9 and 10 have no printed row; 200% of the guideline gives $119,300 and $130,300, where the page's 'for each additional HH member add $11,004' line would give $4 and $8 more; the model follows the 200% rate".
  • Origin of the error. The variable's documentation says "the plan's truncated equivalent SMI percentages or its inconsistent additional-member line", and the entry copies that ambiguous wording. The test comment shows which page the line is on.

The open question has an answer: the plan's percentages change eligibility.

  • The SMI. Maryland's FFY2026 four-person SMI is $154,117 (Case 2).
  • Seven people. 46.00% × 1.35 × $154,117 = $95,706.66, against the guidelines' $97,300. That is a $1,593 gap.
  • Other sizes. At size 1, 39.00% gives $31,254.93 against $31,300.

The year type differs from the code and the sibling entries. The entry uses type: federal-fiscal. The five other MEAP candidates use type: program, and the code reads Maryland's July-June year: smi_rate.yaml and smi_min_size.yaml start 2025-07-01, and md_meap.py says "state FY25 to FY27 (July 2024 to June 2027)".

Effect. The model is not affected. A new entry in a file shared with Axiom misplaces one conflict and merges two distinct ones:

  • (a) the plan's truncated SMI percentages against the guidelines' 200% FPG (up to $1,593);
  • (b) the guidelines' own $11,004 add-on against 200% FPG ($4 and $8 at sizes 9 and 10).

Fix. Rewrite the entry's fields along these lines, keeping its sources and code list:

  years: {from: 2026, to: 2026, type: program, note: program year as the program defines it}
  question: Do Maryland MEAP's FY2026 income limits for sizes 1-10 follow the OHEP guidelines' 200% FPG amounts or
    the state plan's lower SMI percentages?
  readings:
    statute: No statute supplied; the OHEP guidelines print 200% FPG limits for sizes 1-8, add $11,004 per additional
      member, and use 60% of SMI from size 11.
    other: Plan item 2.1 lists 39.00% to 58.70% of SMI for sizes 1-10, each a truncated equivalent of 200% FPG.
  example: Seven people with $96,000 are under the guidelines' $97,300 but over the plan's 46.00% of SMI ($95,706).
  note: PE-US follows 200% of the poverty guideline for sizes 1-10 and 60% of SMI from size 11. For sizes 9 and 10
    it uses 200% of the guideline ($119,300 and $130,300), not the guidelines' $11,004 add-on, which gives $4 and $8
    more (md_meap_income_limit.py, eligibility/fpg_rate.yaml, eligibility/smi_min_size.yaml).

Change the plan citation to "FY2026 LIHEAP State Plan item 2.1, p.9", and update the PR body's description of this candidate. Optionally, a follow-up code PR could reword the variable's documentation so that "its" refers to the guidelines page.

Suggestions

1. Fill the null URLs and drop the stale instructions that the code on main already answers. The PR already did this for three entries ("filled ... from the code's own reference for the same provision"). The same reasoning applies to these entries the PR edits:

  • us-in:rcap-unlicensed-facility-per-diem (lines 2009-2039):
    • Source 1. Use https://www.law.cornell.edu/regulations/indiana/455-IAC-1-3-3, which unlicensed.yaml cites, and name "455 IAC 1-3-3(a)". It is the in-force text, readopted 11/13/2019: "a facility that is not licensed under IC 16-28 ... shall receive per diem reimbursement of twenty-seven dollars ($27)".
    • The note. It can drop "Obtain the exact in-force rule text".
    • The guide. The 2017 guide never prints $37. It gives a county-home income limit of $1,125.41, which is $37 × 365 / 12 = $1,125.42 truncated. The "other" reading could say so.
  • us-mo:wftc-investment-income-boundary (lines 3503-3541):
    • Form. Use https://dor.mo.gov/forms/MO-WFTC_2023.pdf#page=1, which investment_income_limit.yaml cites. Line 3 reads "Do you have investment income equal to or greater than $4,050", and the same form also says "cannot exceed $4,050".
    • FAQ. Use https://dor.mo.gov/faq/taxation/individual/missouri-working-family-tax-credit.html.
    • The note. It can drop "obtain the 2023 form URL".
  • us-ms:liheap-110-percent-poverty-floor (lines 3553-3565):
    • URLs. The three null URLs are in ms_liheap_eligible.py's references: Cornell 8624, the manual (00000693c.pdf#page=58 for the printed table) and MS_Plan_2026.pdf#page=44.
    • The example. It still says "the proposed statutory model awards $650 for propane". That is now main's behaviour (ms_liheap.yaml Case 20), so write "PE-US awards $650".
  • Maryland MEAP manual and plan URLs (lines 2731, 2764, 2767, 2828, 2831):
    • Manual Attachments A and B are file pages 111 and 112 (OHEP-Operations-Manual.pdf#page=111 and #page=112, as unearned.yaml cites).
    • Plan item 1.9 runs over pages 6-7.
  • Statute and regulation URLs the code cites:
    • LA R.S. 47:294 and 47:44.1 (line 2376; amount.yaml, cap.yaml);
    • ME 36 M.R.S. §5219-KK (line 3058; utilities_share_of_rent.yaml);
    • 7 CFR 273.5 (line 3483; per_household.yaml).
  • Optional: the ME TANF citation could name the division: §3762(3)(B)(7-D)(d), and (7-C)(c) for applicants.

2. Fix four page anchors in entries the PR edits. Each predates the PR (#10287), and the code on main already cites the right page.

  • us-mi:household-resources-dependent-amounts (line 3178):
    • The citation says "page 32, line 21", but the link is MI-1040-Book.pdf#page=28.
    • File page 28 is not line 21. File page 32 reads "Line 21: Enter all Social Security, Supplemental Security Income (SSI), and/or Railroad Retirement benefits. Include death benefits and amounts received for minor children or other dependent adults who live with you."
    • household_resources_all_members.yaml links #page=32.
  • us-md:meap-garrett-oil-level-four (plan source):
    • The link is MD_Plan_2026.pdf#page=5, which is items 1.3-1.6.
    • The multiplier is on page 11: "applies a 1.1 multiplier to MEAP benefits for Garrett County residents". garrett_multiplier.yaml links #page=11.
  • us-md:meap-interest-dividends (plan source):
    • The link is #page=5.
    • Item 1.9 starts on page 6, and the "Interest, dividends, or royalties" row is on page 7.
  • us-md:meap-nonqualified-child-household-size (line 2798):
    • The citation says "page 108", but the link is OHEP-Operations-Manual.pdf#page=11, which is the comments section.
    • Page 108 reads "Children under 18 are always counted in the household, regardless of citizenship status", and md_meap_household_size.py links #page=108.
    • Once the link is fixed, the rewritten note can drop "verify the supplied page anchor".

3. Add the deciding file to us-mt:elderly-credit-homestead-costs (lines 3604-3637).

  • The file. mt_elderly_homeowner_or_renter_credit_selected_claimant.py (Limit the Montana elderly homeowner/renter credit to one claimant per household (MCA 15-30-2341(1)) #9900) is where main records this exact conflict: "The statute instead defines both for the homestead (§ 15-30-2337(5), (10)), which would give the one claimant the household's whole property tax and rent; that reading is not modeled."
  • Why it decides the entry. It also picks the claimant whose credit is largest.
  • The current list. ..._pre_one_claimant.py is listed, but its comments don't touch the question.

4. us-ky:ktap-payment-scale-household-size: add the regulation and note the web page's audience.

  • The regulation. 921 KAR 2:016 §9(2)(a), the version effective 2-16-2023, labels its table "Number of Eligible Persons", like manual MS 2820. Adding it as a regulation source makes the KY FACES label the only outlier.
  • The page's audience. The KY FACES page sits in the Kinship section and is written for relative caregivers. In a child-only case the eligible persons are the children, so "Number of Children" may describe that audience rather than a different rule. The readings or note could say so.
  • The code's count. It keys on spm_unit_size, capped at 7. That is a proxy for the eligible-person count; MS 2810 says "Do not count technically excluded/sanctioned members". The note's "keyed to assistance unit size" could say "proxied by SPM unit size".

5. Point us-mn:msa-guardian-fee-cap at its code on main (lines 3250-3274).

  • The entry. It has code: [] and prs: [], so it fell outside the PR's "a PR each one names has since merged" filter.
  • The code. It has been on main since Add Minnesota Supplemental Aid (MSA) #8215 (merged 2026-06-01) and follows the statute, as the entry records: mn_msa_guardian_fee.py applies min_(unit_gross * p.rate, p.max_amount), with guardian_fee/rate.yaml at 0.05 and max_amount.yaml at 100.
  • Fix. List those three files and Add Minnesota Supplemental Aid (MSA) #8215. The example still holds: 5% × $994 = $49.70, and 5% × $1,029 = $51.45.

6. One reason in the PR body's unknown list doesn't fit its entry.

  • The body. It lists us-md:premium-assistance-200-fpl-endpoint as "(no source is a statute)".
  • The entry. Its first source cites "Md. Insurance Article §31-125; COMAR 14.35.21", and its statute reading is "The statutory description does not specify the endpoint".
  • The header's definition. It defines gap-fill as "the statute is silent and the code follows the form".
  • Fix. Either set gap-fill, since PE-US fills the silence with the Board design and the MIA exhibit, or keep unknown and give the reason that the Board-design source is not a form.

7. Optional: one older Maryland conflict outside the refreshed PRs.

  • The parameter. md/msde/ccs/copay/weekly_amount.yaml follows the flat $1/$2/$3 schedule from May 2022, which Chapter 717 of 2024 caps at January 1, 2024 levels.
  • The conflict. It cites "COMAR 13A.14.06.12 - codifies the historical regional matrix (superseded operationally)". That is a regulation against agency practice and statute, and it has no candidate.
  • Scope. It isn't from any of the 24 PRs, so a follow-up would do.

Verified correct

Each refreshed entry, against main:

Candidate What the code on main does follows
us-in:eitc-current-versus-frozen-federal-base In the static-conformity branch, in_eitc.py returns eitc × match_rate; its comment argues the post-snapshot changes are "inflation indexing only". in_eitc_eligible.py tests the frozen IRC: "2026-01-01" if period.start.year >= 2026 else "2023-01-01" conflicting-source
us-in:rcap-unlicensed-facility-per-diem unlicensed.yaml: 27 from 2004-01-01, 37 from 2008-07-01. in_ssp_rcap.py multiplies it by 365/12. #7887 added both unknown
us-ks:medicaid-parent-dependent-child-age age_limit.yaml KS: 19 from 2014; student_option.yaml KS: false, with the S25 guard comment unknown
us-ks:snap-crop-share-farm-rent-classification snap_farm_rent_income.py = max_(farm_rent_income, 0), citing 7 U.S.C. 2014(d)(9). It is listed in unearned.yaml, whose comment names KS memo 99-10-05 unknown
us-ky:ktap-payment-maximum-2025 payment_maximum.yaml: 242-627 from 2025-11-01, with the comment that 921 KAR 2:016 "still shows the 2023 ($372-$964) schedule". #8942 added that comment unknown
us-la:cpi-u-indexing-method 2026 = 12,838 / 25,676 / 12,324. get_la_cpi_u_percentage_increase uses the December-over-December NSA change rounded to one decimal. extend_la_cpi_u_indexed_amounts doubles the rounded single amount for JOINT, HOH and SURVIVING_SPOUSE gap-fill
us-ma:circuit-breaker-snap-income snap and wic are commented out of disallowed_deductions.yaml, under the DOR-versus-item-2 comment. ma_scb_total_income.py adds the list for the head and spouse statute
us-ma:connectorcare-income-bands-2026 fpl_floor 1 and fpl_limit 4 from 2026. ma_connector_care_eligible.py applies >= floor and <= limit. Both parameter comments record Plan Types 1 and 3D unknown
us-md:meap-garrett-oil-level-four md_meap.py: np.floor(amount × 1.1 + 0.5) for Garrett. Its documentation records the FY27 table's "$1,210 where 1.1 times $980 is $1,078" unknown
us-md:meap-interest-dividends, -lump-sums-americorps, -strike-pay unearned.yaml includes interest_income, ordinary_dividend_income and strike_benefits and omits lump sums. The md_meap_countable_income.py documentation records each departure unknown
us-md:meap-nonqualified-child-household-size md_meap_household_size.py sums qualified | child (age < 18) unknown
us-md:premium-assistance-200-fpl-endpoint initial.yaml and final.yaml: the 150-200% band ends at 0.02, and the 200-250% band starts at 0.02. The comment says HGO p.58 prints 1.00% twice unknown (see Suggestion 6)
us-me:property-tax-credit-unknown-utilities countable_rent.py: when the amount is 0 (unknown), it subtracts rent × utilities_share_of_rent (0.15 from 2014). The rate's comment says "the statute contains no unknown-amount estimate" gap-fill
us-me:sales-tax-fairness-credit-mixed-joint-return ~every_filer_dependent & ~separate; the comment says neither 5213-A(6)(C) nor subsection 2 settles the mixed case gap-fill
us-me:tanf-incapacitated-adult-care-deduction care_recipient = is_child | incapacitated_adult, with adults on the age-2+ tier ($175) conflicting-source
us-mi:household-resources-dependent-amounts mi_household_resources.py sums sources over the head and spouse (tax_unit_non_dep_add), except household_resources_all_members.yaml (lines 21, 22 and 27), which it sums over every member conflicting-source
us-mo:medicaid-published-dollar-limit-rounding medicaid_income_eligible takes ceil(round(fpg × limit / 12, 2)) and admits income at or below it in every state. MHABD uses np.ceil(limit_pct × fpg / 12); MHF uses the published schedule gap-fill
us-mo:property-tax-credit-rounding mo_property_tax_credit.py takes np.ceil of the phase-out; the comment says the published charts round the credit down while the statute says nearest dollar conflicting-source
us-mo:snap-student-child-care-claim-cap per_household.yaml (MO true from 2025; the comment calls the start "a modeling choice"). meets_snap_parent_exception.py gives gets_claim to rank 0 only in capped states gap-fill
us-mo:wftc-investment-income-boundary investment_income <= p.investment_income_limit in every year (4,050 / 4,300 / 4,400) statute
us-ms:liheap-110-percent-poverty-floor (income <= limit) | (income < floor), with the floor at 1.1 × the lagged guideline, rounded to cents (#10279) statute (changed, correct)
us-mt:elderly-credit-homestead-costs property_tax_billed adds the tax unit's real_estate_taxes. gross_rent is max_(rent, household tax if the unit is billed nothing) conflicting-source
us-mt:elderly-credit-nonloss-agi-adjustments The sources start from adjusted_gross_income_person and add back loss_ald and lost capital gains statute
us-mt:elderly-credit-refundable-credit-timing federal_refundable_credits = head × income_tax_refundable_credits (current return). mt_elderly_homeowner_or_renter_credit_received is an input unknown
us-mt:elderly-credit-rent-free-occupants As in the homestead row: the household's property tax counts as rent when the unit is billed nothing, taking the larger of the two conflicting-source
us-mt:medical-floor-agi-base Through 2023 the floor uses mt_agi_joint; from 2024 it uses the federal medical_expense_deduction. mt_standard_deduction_joint also uses mt_agi_joint gap-fill

The new KY candidate. payment_maximum.yaml's comment records both axis labels, as the entry says. See Suggestion 4.

The PR body's other claims:

Sources I read (generic User-Agent):

  • 455 IAC 1-3-3(a) (Cornell): $27 unlicensed, $39.35 licensed.
  • The RCAP Provider Guide (September 7, 2017): $1,125.41 county-home limit, with no per diem printed.
  • 921 KAR 2:016 §9(2)(a) (effective 2-16-2023): "Number of Eligible Persons", $372-$964.
  • KY DCBS manual MS 2820 (OMTL-682, R. 11/1/25): "Number of Eligible Persons", $242/$293/$341/$426/$498/$562/$627.
  • KY FACES: "Number of Children", $242-$627.
  • 22 M.R.S. §3762(3)(B)(7-D)(d): "All actual child care costs necessary for work, except that the department may limit the child care disregard to $175 per month per child or $200 per month per child under 2 years of age or with special needs".
  • MD FY2026 plan: pages 5-7, 9 and 11.
  • OHEP Income Guidelines FY2026.
  • OHEP Operations Manual: pages 11, 108, 111 and 112.
  • 2023 Form MO-WFTC: "equal to or greater than $4,050".
  • 2025 MI-1040 book: file pages 26-33.
  • ConnectorCare Overview 2026, page 1: "between 100% and 400%"; Plan Types 2A-3C.
  • 956 CMR 12.04, page 6: "at or below 500 percent"; Plan Type 1 below 100%; 3D at 400-500%.
  • LA RIB 26-019: 2.7%; $12,838 / $25,676 / $12,324.
  • 2025 MT Form 2 instructions, page 47: "the property taxes billed for that property can qualify as rent only".
  • MA FY2026 plan, page 9: 60.00% SMI for all sizes.
  • Indiana PY2026 EAP manual, page 23: 60% SMI.

Coverage against main's comments and our recent reviews:

Hand checks (12):

  1. MS floor example:
    • Floor: 1.1 × (15,650 + 10 × 5,500) = 77,715.
    • Eligibility: $77,000 is above the $75,593 limit and below the floor, so the household is eligible.
    • Payment: the propane column's open fourth band gives 800 − 3 × 50 = 650 (Case 20).
  2. KY schedules:
    • Manual: 242 / 293 / 341 / 426 / 498 / 562 / 627 equal the 2025-11-01 parameter values.
    • Regulation: 372 / 450 / 524 / 656 / 766 / 864 / 964 equal the 2023-02-16 values.
  3. LA:
    • Single: 12,500 × 1.027 = 12,837.50, rounded to 12,838.
    • Joint: 2 × 12,838 = 25,676; adjusting 25,000 separately gives 25,675.
    • Unrounded change: 12,500 × 324.054 / 315.605 = 12,834.6, rounded to 12,835.
    • Retirement: 12,000 × 1.027 = 12,324.
  4. IN RCAP: 37 × 365 / 12 = 1,125.42, which is the guide's $1,125.41 truncated. 27 × 365 / 12 = 821.25.
  5. MD Garrett: floor(1.1 × 980 + 0.5) = 1,078.
  6. MD premium assistance: $53 × 12 / $31,300 = 2.03%; 1% × 31,300 / 12 = $26.08.
  7. MO WFTC: 4,050 ≤ 4,050, so eligible; 4,051 is not.
  8. MN guardian fee: 5% × $994 (2026 FBR) = $49.70; 5% × $1,029 (2027) = $51.45.
  9. IN floor: 1.1 × (15,650 + 15 × 5,500) = 107,965, above the 104,543 limit for 16. For 15, the limit is 102,607 and the floor 101,915, so the limit still applies.
  10. MD floor:
    • Size 29: 1.1 × (15,650 + 28 × 5,500) = 186,615, against floor(0.6 × 154,117) × 2.01 = 185,864.
    • Size 28: the limit is 183,090, above the 180,565 floor.
  11. MA floor: 1.1 × (15,650 + 33 × 5,500) = 216,865, against 99,573 × 2.16 = 215,077.68 for 34 people. Size 33 gives 212,090.49, above the 210,815 floor.
  12. MD plan percentages:
    • Size 1: 31,300 / (0.52 × 154,117) = 39.06%, printed as 39.00%.
    • Size 7: 97,300 / (1.35 × 154,117) = 46.77%, printed as 46.00%, which gives $95,706.66.
    • Size 9: 119,300 / (1.41 × 154,117) = 54.899%, printed as 54.89%.
    • Guidelines add-on: 108,300 + 11,004 = 119,304, against 119,300.

Static run of test_source_conflicts.py's candidate rules, at the head (225 candidates) and on the merged tree (227):

  • What I checked: unique, well-formed ids in sorted order; none in the vendored register; required keys; valid state, follows, source kind and years, with a note for program years; at least two sources; https or null URLs; code paths tracked at that tree; PR URLs; the privacy patterns, line by line and on parsed values.
  • Result: no problems.
  • Code paths on main. Every head code path is also tracked on main.
  • The ratchet. test_conflict_sounding_comments_do_not_grow is unaffected: the PR touches no file under variables/, parameters/ or reforms/.

Housekeeping:

  • Changelog. changelog.d/source-conflict-candidates-in-mt.changed.md is top level and typed changed.
  • Partner tests. No partner test file is touched.

Axiom line assessment

Valid. The body has axiom: n/a: source-conflict backlog metadata only; no model change. The PR changes only backlog metadata and a changelog fragment, with no variable, parameter or test, so a housekeeping n/a is the right form. No rulespec-us issue is needed.

CI status

27 pass, 0 pending, 0 failed. This is run 39627529795 on head df1b6ece99 (pull_request, 2026-10-21 15:38 to 20:21 UTC). I re-checked gh pr checks 10351 at the end of this review.

  • The source-conflict tests. Rest (Python, shard-1 + variables) ran test_source_conflicts.py: 103 passed and 0 failed, matching the PR body. The passing tests include:
    • test_candidate_fields;
    • test_candidates_have_unique_well_formed_ids;
    • test_candidates_are_not_yet_registered;
    • test_candidates_are_public;
    • test_conflict_sounding_comments_do_not_grow.
  • Known main-caused ratchet failure. It doesn't apply: the run started after 13:24 UTC on 2026-10-21.
  • Also passed: all 11 Baseline shards, all 9 Contrib shards, Household API Partners, Microsimulation, both Rest jobs, PR validation, Package and compatibility, and Quick Feedback.
  • Before the next push, merge main. The head's uv.lock is lock revision 3. main moved to revision 5 with Stop trusting the compromised eMudhra CA certificates #10377 on 2026-10-22. On Point the FL and SD LIHEAP conflict candidates at their code on main #10349, a run after Stop trusting the compromised eMudhra CA certificates #10377 failed "Package and compatibility" for exactly this reason. Merging main avoids it, and Should 1 needs main anyway.
  • Locally: no tests or models were run. I parsed the candidates YAML only.

Branch status

Verdict: REQUEST_CHANGES

@DTrim99 DTrim99 left a comment

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

All 67 code paths exist on main and match their notes, the follows values are right, and the candidate rules pass at the head and on the merged tree. Two should-address items: (1) add 110%-floor candidates for MA LIHEAP (#10279), IN EAP and MD MEAP (#10281), which main records in comments (the same gap as FL on #10349); (2) the new us-md:meap-income-limit-plan-percentages entry misplaces the additional-member line (it's on the OHEP guidelines page, not the plan), and the year type should be program. Seven suggestions, mostly URLs and anchors. Full report in the comment above.

MaxGhenis and others added 2 commits October 24, 2026 11:59
Main's #10441 moved some of the edited candidates into the shared register,
so their blocks go and the edits to them are dropped; #10299 rewrote the
three SNAP timing candidates, so main's text stands for those.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
…ates

Refresh entries from code merged since the first round, add the missing
110%-floor candidates for Indiana, Massachusetts and Maryland, and fix
the page anchors and citations the reviews flagged.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@MaxGhenis

Copy link
Copy Markdown
Contributor Author

Thanks, @DTrim99. Addressed at 7bc14cd, after merging main (which also brings uv.lock up to date):

Should 1 (floor candidates). Added us-in:liheap-income-floor-vs-smi-limit, us-ma:liheap-income-floor-vs-smi-limit and us-md:meap-income-floor-vs-smi-limit, using your drafts. Each example figure was recomputed from the code's formula and parameters on main.

Should 2 (us-md:meap-income-limit-plan-percentages). Took your rewrite: the question, both readings, the note and the example. The plan citation is now "FY2026 LIHEAP State Plan item 2.1, p.9". Its years now use program, like the other MEAP entries, instead of federal-fiscal; the Sol review flagged the same field.

Suggestion 1 (null URLs and stale instructions).

  • RCAP: the source is now 455 IAC 1-3-3(a) at Cornell. The note records that the guide's $1,125.41 is $37 a day and drops "Obtain the exact in-force rule text".
  • MO WFTC: added the form and FAQ URLs, and the note drops "obtain the 2023 form URL".
  • MS floor: filled the three URLs, and the example now reads "PE-US awards $650".
  • MD MEAP: the manual URLs point to pages 111 and 112, and the plan URLs to pages 6 and 7.
  • LA, ME and 7 CFR 273.5 URLs are filled, and the ME TANF citation names (7-D)(d).

Suggestion 2 (anchors). All four are fixed:

  • MI MI-1040 book: #page=32.
  • Garrett multiplier: #page=11.
  • MD interest and dividends: #page=7.
  • MD nonqualified child: #page=108. That note no longer asks to verify the anchor.

Suggestion 3 (MT homestead). The code list now has mt_elderly_homeowner_or_renter_credit_selected_claimant.py in place of the pre-one-claimant file.

Suggestion 4 (KY K-TAP scale). Added 921 KAR 2:016 §9(2)(a) as a regulation source. The note now says the scale is proxied by SPM unit size capped at seven, and that the KY FACES page is written for kinship caregivers.

Suggestion 5 (MN guardian fee). #10441 moved this entry to the shared register as msa-guardian-fee-cap, with the code paths you list, so it is no longer a candidate here.

Suggestion 6 (body reason). us-md:premium-assistance-200-fpl-endpoint stays unknown. The body now gives the reason that the statute is silent and PE-US fills it from the Board design and the MIA exhibit, which are not a form, so gap-fill as the header defines it does not fit.

Suggestion 7 (MD CCS copay). Out of scope here, since it isn't from a refreshed PR. I've left it for a follow-up.

Also changed.

The guard test passes (103) at the new head.

MaxGhenis and others added 2 commits October 24, 2026 14:12
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Brings in #10350, #10354 and #10444. Git merged candidates.yaml cleanly,
and the result equals main's file with this branch's 29 refreshed and 5
added candidates reapplied. #10354 changed only
us-in:claim-of-right-federal-deduction-eligibility, which this branch
does not edit.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@MaxGhenis
MaxGhenis dismissed DTrim99’s stale review October 25, 2026 11:41

Dismissed at @MaxGhenis's direction. Each point in this review is fixed or answered in #10351 (comment), and independent reviews approved the current head. Happy to take up anything still open in a follow-up.

@MaxGhenis
MaxGhenis merged commit e8dbb6c into main Oct 25, 2026
32 checks passed
@MaxGhenis
MaxGhenis deleted the source-conflict-candidates-in-mt branch October 25, 2026 11:43
@MaxGhenis

Copy link
Copy Markdown
Contributor Author

Merged at the reviewed head 777ab87 with --admin. The only block was the required approving review.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants