# Verification record — `pk/fbr/annual-individual-income-tax-return-it-2` v1.0.0 This file is the **source-review record** for this document version, per the [`manual-source-review-v1`](../../../../../practices/manual-source-review-v1.md) practice. ## Current claim - **`status`:** `draft` - **`verification.method`:** `manual-source-review-v1` - **`verification.lastVerifiedAt`:** `2026-07-31` ## Why this schema and why now (GOV-5738, "GovSchema Standard Research") This cycle re-scanned `CATALOG.md` fresh rather than trusting the in-repo state from memory. Every jurisdiction currently disclosed at 5 of 6 verticals was re-checked against the document's own vertical-history entries: **AE** (Passport), **BA** (Visa), **BR** (Visa), **CZ** (Passport), **GR** (National ID), **ID** (National ID), **JM** (National ID), **MK** (Visa), **MN** (National ID), **MT** (Taxes), **NO** (Passport), **PL** (Visa), **RW** (Taxes), **SK** (Visa), **TN** (DMV), **TT** (National ID) — 16 in total. Every one of these gaps was independently confirmed, from the document's own prior-cycle text, to already be a re-screened dead end (in-person/biometric-only, login/CAPTCHA-gated, or a confirmed duplicate of an already-modelled EU-harmonized template) rather than a genuinely new, untested lead — matching the immediately preceding GOV-5731 cycle's own finding for a largely overlapping set. With no jurisdiction-closing candidate available, the strongest actually-open, pre-sourced candidate was this worksheet: both the sibling `pk/fbr/annual-individual-income-tax-return-it-1b` document's own v1.0.0 (GOV-3104, 2026-07-15) and the `pk/fbr/wealth-statement` document's own v1.0.0 (GOV-5731, 2026-07-31) explicitly named the 'IND (BUS PLUS)' worksheet (Form IT-2) as open backlog. This cycle picked up that pre-flagged companion rather than starting a new jurisdiction from scratch. **Aside, disclosed for the record:** a separate, non-schema markdown document (`GovSchema_Research_Status_2026-07-31.md`, merged directly by a human recovering a crashed background agent's commit, PR #758) was found in the repository root this cycle, claiming Brazil is "missing from ALL 6 verticals." This is independently contradicted by both `CATALOG.md`'s own `## By Jurisdiction` table (`BR` — 6 schemas, 5 of 6 verticals: Passport, DMV, Business Formation, Taxes, National ID all ✓, only Visa a confirmed dead end) and the registry contents on disk (`registry/br/pf`, `registry/br/sp/jucesp`, `registry/br/mg/detran`, `registry/br/rfb`, `registry/br/tse`, `registry/br/pr/iipr`). The PR's own description already flags this and other figures in that document as stale and directs readers to treat `CATALOG.md`, not that document, as the source of truth — no further correction was needed this cycle. ## Sources examined - **Document `(id, version)`:** `pk/fbr/annual-individual-income-tax-return-it-2` / `1.0.0` - **Spec version:** GovSchema `0.3.0` - **Authority:** Federal Board of Revenue (FBR). - **Primary source:** - Direct `.xlsx` (the same workbook already modelled twice, as `annual-individual-income-tax-return-it-1b` and `wealth-statement`): — independently re-fetched this cycle via plain `curl`: HTTP 200, **size 104,097 bytes**, sha256 `4e2f9874e9a910713fae9a182fc5c5a578bcfd2649d2031f8144c8b5fba96b9f` — byte-identical to the figures both the GOV-3104 (2026-07-15) and GOV-5731 (2026-07-31) cycles recorded for this same workbook, confirming no revision. No login/CAPTCHA gate on this direct `download1.fbr.gov.pk` URL. - Native Office Open XML `.xlsx` (a zip archive of XML parts), not a PDF. No `unzip`, `pip`, or `openpyxl` available in this environment; unzipped directly with Python's built-in `zipfile` module. The worksheet `IND (BUS PLUS)` (Form IT-2, this document's scope) is confirmed via `xl/workbook.xml`'s own `` element (``), mapped to `xl/worksheets/sheet2.xml` via `xl/_rels/workbook.xml.rels`'s `rId2` relationship, and cross-checked against the sheet's own printed title row (" RETURN OF TOTAL INCOME / STATEMENT OF FINAL TAXATION UNDER THE INCOME TAX ORDINANCE, 2001 (IT-2)"). - Parsed `sheet2.xml`'s `sheetData` rows (row 1 through row 170, the sheet's last populated row, spanning two printed pages within the one worksheet), `xl/sharedStrings.xml`'s shared string table, and the sheet's own `dataValidations` block (4 rules) cell-by-cell from the raw XML. ## Scope decisions The `IND (BUS PLUS)` worksheet is structurally the business-income sibling of the already-modelled `IND (PROP-CG-OS)` worksheet (Form IT-1B): the same identification block, the same headline-income-line shape, and a nearly identical verification/declaration block, but for individuals deriving income under the head Business (plus any head other than Salary) rather than Property/Capital Gains/Other Sources. **In scope, at the level the sheet's own `dataValidations` confirm is directly fillable, mirroring IT-1B's own scope boundary exactly where the two sheets share structure:** - Identification: `taxpayerName`, `taxpayerCnic`, `taxpayerNtn`, `taxpayerAddress`, `taxYear`. - Sr. 11–13: `businessIncome`, `capitalGainsIncome`, `otherSourcesIncome` (the Sr. 13 headline total). - Sr. 14–25 (Other Sources sub-items): `otherSourcesReceipts`, `royaltyIncome`, `profitOnDebt`, `profitOnDebtExceeding36Million`, `groundRent`, `subLeaseRent`, `leaseRentWithPlantAndMachinery`, `annuityPension`, `otherReceipts`, `otherSourcesDeductions`, `accountingDepreciation`, `otherDeductions`. - Sr. 26–30 (Foreign Income and its four sub-categories): `foreignIncome`, `foreignBusinessIncomeLoss`, `foreignPropertyIncomeLoss`, `foreignCapitalGainIncomeLoss`, `foreignOtherSourcesIncomeLoss`. - Sr. 31–33: `aopUntaxedIncomeShare`, `aopTaxedIncomeShare`, `totalIncome`. - Sr. 34–38 (Deductible Allowances and its three components, plus Taxable Income): `deductibleAllowancesTotal`, `zakatDeduction`, `workersWelfareFundDeduction`, `educationExpensesDeduction`, `taxableIncome`. - The verification/declaration block: `declarantName`, `declarantCnic`, `declarantCapacity`, `declarationDate`. **A genuine structural difference from IT-1B, not a scope narrowing:** the IT-1B sheet has no Other Sources sub-item breakdown at all (its own Sr. 12 "Income / (Loss) from Other Sources" is a single line with no feeding rows) and no Foreign Income sub-categories (its own Sr. 13 "Foreign Income" is likewise a single line). This sheet's own rows 21–37 print twelve individually labelled Other Sources sub-items and four individually labelled Foreign Income sub-categories, each validated fillable by the sheet's own `dataValidations` (decoded below). These are modelled individually — the same registry convention already used for Deductible Allowances (a headline total plus its named components, present in both this document and IT-1B) and for Wealth Statement's own Inflows/Outflows exception — rather than collapsed into their respective headline totals alone, since each is a fixed, small, uniquely labelled line item, not a repeating table shape (which would require GSP-0009, still Proposed and CEO-gated). **Out of scope, mirroring IT-1B's own scope boundary exactly:** - The Property income sub-item breakdown (Sr. 1–10, rows 8–17). Row 8's own F8/G8/H8 cells carry **no** data-validation rule at all — the identical computed-subtotal signal IT-1B's own Sr. 1 (row 9) carried, confirming Property remains a computed subtotal on this sheet too, with no direct-entry validation of its own. - The workbook's own per-line three-column split (`F`/`G`/`H` — "Total Amount" / "Amount Exempt from Tax or Subject to Fixed/Final Tax" / "Amount Subject to Normal Tax"). Not modelled reliably: the sheet's own data-validation ranges cover the three columns inconsistently. For example, row 18 (Sr. 11, Business Income) validates all three columns (`F18:H18`), but rows 33–37 (Sr. 26–30, the Foreign Income total and its four sub-categories) validate only `F` and `H` — the `G` column is **not** validated for any of these five rows (`G21:G32`, the sheet's only decimal-typed validation rule, stops at row 32, one row short of Sr. 26 at row 33). This is the same class of per-row column-validation inconsistency IT-1B's own verification record already documented for its own sheet, not fabricated or corrected here — this document reports a single headline figure per line, as IT-1B does. - The Computations/tax-liability section (Sr. 39–62, rows 47–70): tax chargeable, normal income tax, tax credits, four distinct minimum-tax adjustment lines, six turnover-tax lines at named statutory percentages (0.25% through 1.5% under section 113), tax on high-earning persons (u/s 4C), tax on deemed income (u/s 7E), advance/admitted/refundable/ demanded income tax, a refund-adjustment line, Workers' Welfare Fund, and agriculture income/tax — arithmetic derived from the return, not primary taxpayer-supplied data, the same exclusion basis IT-1B applied to its own Sr. 21–33. - The Final/Fixed/Minimum/Average/Relevant/Reduced Tax withholding schedule — a second printed page (rows 74–169, confirmed by the worksheet's own page-2 header at row 74 and a "2/2" page-number cell), roughly 150 individually named withholding-tax-rate categories spanning sections 148 (imports), 150 (dividends), 152 (non-resident payments), 153–156 (goods/services/contracts/prizes), 233–236Z (brokerage, utilities, tolls, petroleum, bonus shares, capital gains on immovable property and securities at named holding-period bands and rates), and 156A/233A (commission/brokerage). All confirmed fillable by the sheet's own `dataValidations` (`F80:G146`, `H80:H169`, `F148:G169` — unlike the Property subtotal, these rows genuinely are data-entry cells, not computed), but excluded on the same disproportionate-expansion basis IT-1B's own verification record already applied to this identical schedule (there, ~30 rows counted from IT-1B's own sheet; here, roughly five times as many) — a schedule of this size would require either GSP-0009 (repeating-value fields, still Proposed and CEO-gated) or flattening ~150 distinct field names for a single companion-schema cycle, a disproportionate expansion relative to the value delivered. - Annexes A through F, which remain this workbook's own disclosed, cross-schema backlog (shared with the sibling IT-1B and Wealth Statement documents). ## Disclosed structural findings 1. **`NTN` is required here, unlike the sibling IT-1B document.** This sheet's own row 5 label reads `"NTN*"` (with an asterisk), while IT-1B's own row 5 reads `"NTN"` (unstarred, and modelled `required: false`). The underlying data-validation rule on the NTN cell (`H5`, whole number 1–99,999,999) is otherwise identical in shape to IT-1B's own `J5` rule. `taxpayerNtn` is modelled `required: true` here, a genuine and disclosed difference driven by the source form's own printed marking, not a judgment call. 2. **Tax Year internal inconsistency.** The workbook's own title (`Individual Paper Return for Tax Year 2024`) and this sheet's own second-page repeated header (row 76, cell `H76`, pre-filled `2024`) both read 2024, but this same sheet's own first-page header (row 4, cell `H4`) is itself pre-filled the literal numeric value `2023` — a genuine internal inconsistency in the source spreadsheet. Disclosed here and in the `taxYear` field's own description; not corrected by this schema. (The sibling IT-1B sheet's own row 4 is consistently pre-filled `2024` with no equivalent defect.) 3. **Verification block reassembled from three separate text fragments, not one paragraph.** Unlike IT-1B's own verification block (a single contiguous cell range at rows 76–77), this sheet's declaration spans row 71 (`"Verification"`, a blank name cell, `", CNIC No."`, the `F71:G71` CNIC input cells, `", in my capacity"`, a blank capacity cell), row 72 (the long `"as Self / Representative (as defined in section 172 of the Income Tax Ordinance, 2001) of the Taxpayer named above, do solemnly declare..."` continuation, a separate shared-string entry), and row 73 (`"Signature:"` / `"Date:"`). Reassembled verbatim into `documents[].statement`; the underlying legal text is otherwise identical in substance to IT-1B's own. 4. **Foreign Income sub-category `G`-column validation gap.** As noted under "Scope decisions" above: Sr. 26 (Foreign Income total, row 33) and its four sub-categories (Sr. 27–30, rows 34–37) are validated fillable on columns `F` and `H` but not `G` — the sheet's only decimal-typed validation rule (`G21:G32`) stops one row short. Disclosed as a real, observed data-validation gap, not a fabricated or corrected value. 5. **A likely source-document typo, quoted verbatim and disclosed, not corrected.** Sr. 27's own printed label reads `"Foreign Bussiness Income/Loss"` (double "s" in "Bussiness") — quoted exactly as printed in the field's `sourceRef`, with `[sic]` noted, rather than silently corrected to "Business." 6. **CNIC constraint re-confirmed identical across all three documents from this workbook.** The sheet's own `dataValidations` block carries one `type="whole"` rule spanning both `C5` (taxpayer CNIC) and `F71:G71` (declarant CNIC), bounded `[1000000000000, 9999999999999]` — the identical 13-digit constraint already modelled on both `annual-individual-income-tax-return-it-1b` and `wealth-statement`, confirming the same NADRA CNIC format applies uniformly across every worksheet of this workbook. ## Conformance fixtures (Phase 3) 7 fixtures committed under `conformance/pk/fbr/annual-individual-income-tax-return-it-2/1.0.0/`: 2 valid scenarios plus 5 mutation-control fixtures, each derived from one of the valid fixtures by a single targeted mutation. All 7 were run against a from-scratch, ephemeral field-by-field conformance checker (derived directly from this schema's own `fields[]`/`documents[]`, not committed to the repo) before being finalized: - `valid-self-filer-business-income-only.json` (a self-filing business individual reporting only Business Income and Total/Taxable Income, no Other Sources/Foreign/AOP detail) — **0 errors**. - `valid-representative-filed-full-breakdown.json` (a full Other Sources sub-line breakdown, a Foreign Income sub-category, both AOP share types, and all three deductible-allowance components, filed via a representative) — **0 errors**. - `mutation-control-missing-required-field.json` (drops `taxpayerAddress`) — **exactly 1 error**. - `mutation-control-missing-required-ntn.json` (drops `taxpayerNtn` — statically required here, unlike the sibling IT-1B document) — **exactly 1 error**. - `mutation-control-invalid-cnic-pattern.json` (sets `taxpayerCnic` to an 11-digit value) — **exactly 1 error**. - `mutation-control-negative-amount.json` (sets `businessIncome` to `-3200000`, violating `minimum: 0`) — **exactly 1 error**. - `mutation-control-invalid-enum-declarant-capacity.json` (sets `declarantCapacity` to `GUARDIAN`, not in the enum) — **exactly 1 error**. ## Structural validation - `node tools/validate.mjs` — **695/695** (full registry, this document included). - `node tools/validate-ajv.mjs` (ajv 2020-12 against `spec/v0.3`) — **695/695**. - `node tools/verify-sources.mjs registry/pk/fbr/annual-individual-income-tax-return-it-2/1.0.0` — 1 directory, 3 URLs checked, **0 warnings**, **0 failures**. - `npm run build-index` re-run in `tools/govschema-client/` to regenerate `registry-index.json` with this document included (695 entries). ## Maturity `structural-reference`: the source workbook's own printed "IND (BUS PLUS)" (Form IT-2) structure — identification block, the headline income lines and their Other Sources/Foreign Income sub-categories, the Deductible Allowances total and its three components, and the verification declaration — is fully transcribed from the genuine, currently-served official Tax Year 2024 edition (a fillable Excel workbook, not a government online-filing system), but no live filing through FBR's own IRIS e-filing channel was attempted. GovSchema is an independent, non-profit standards body and is not affiliated with, endorsed by, or operated by the Islamic Republic of Pakistan or the Federal Board of Revenue.