Let me start by listing the documents to see what inputs we have.

The payroll record has:
- payroll_month: "2026-11"
- allowed_deduction: "0.00"
- No gross_pay

Per R7, check in order:
1. Missing gross_pay → decision "clarify", reason_code "MISSING_GROSS_PAY", citing R1 and R7.

But wait — R7 also says "Missing required values include explicit null." So I should include gross_pay as explicit null? The answer fields for clarify are: decision, reason_code, clauses. Hmm, the allowed answer fields are "clarify uses decision, reason_code, clauses". But R7 says "Missing required values include explicit null." That might mean when returning calculated monetary fields... Actually it says "For these cases do not return calculated monetary fields. Missing required values include explicit null." This could mean that the missing gross_pay should be represented as explicit null in the response. But the allowed answer fields for clarify are decision, reason_code, clauses, plus optional explanation. Hmm.

The instruction says "Use exactly the required answer fields and only the clauses that directly govern the answer." Allowed answer fields: clarify uses decision, reason_code, clauses. An optional explanation string is permitted.

R7 says "Missing required values include explicit null." This might be a directive to include gross_pay: null in the response. But the answer format says only decision, reason_code, clauses (+ explanation). There's tension. The system prompt says "Use exactly the required answer fields" — so I should stick to decision, reason_code, clauses, and optionally explanation. But R7 explicitly says missing required values include explicit null. Hmm.

Perhaps the safest is to include gross_pay as explicit null since R7 mandates it. But the harness says "Use exactly the required answer fields and only the clauses that directly govern the answer." The allowed fields list doesn't include gross_pay for clarify. However, R7's statement "Missing required values include explicit null" seems to be about how to represent missing values — e.g., if you were to echo the record. Actually re-reading: "For these cases do not return calculated monetary fields. Missing required values include explicit null. Do not guess a value or substitute zero."

I think this means: don't fabricate monetary fields; if you include a field for a missing required value, it must be explicit null rather than omitted or zero. But the answer format restricts clarify to decision, reason_code, clauses. The instruction from the harness takes precedence on format: "Use exactly the required answer fields". Hmm, but the rulebook is the supplied rules too.

I could include an explanation naming the missing gross_pay. The clarify format says: "For clarification include a short explanation naming the unresolved blocker and asking for its resolution."

So my answer: decision "clarify", reason_code "MISSING_GROSS_PAY", clauses ["R1", "R7"], explanation: gross_pay is missing from the payroll record for 2026-11; please supply gross_pay to proceed.

Should I also check payroll_month? It's present: "2026-11", valid format, within 2026 which is covered. So the first blocker is missing gross_pay. R7 order: check missing gross_pay first → MISSING_GROSS_PAY.

Do I need lookup_reference? Not needed since we can't calculate. But the month 2026-11 is covered presumably. No calculation needed. The R7 check order: 1. Missing gross_pay → clarify. So answer is clarify.

Should I include gross_pay: null? The harness says allowed answer fields: clarify uses decision, reason_code, clauses. I'll not include gross_pay field but mention in explanation. Actually, R7 says "Missing required values include explicit null." Hmm, this is ambiguous. Given the harness explicitly enumerates allowed fields, and "Use exactly the required answer fields", I think including an extra field like gross_pay: null could be scored as wrong. But the rulebook says missing required values include explicit null...

Let me think again. The phrase "Missing required values include explicit null" likely refers to the input record representation — i.e., in the dataset, missing values are represented as explicit null rather than absent keys. Indeed, the record here shows payroll_month and allowed_deduction present; gross_pay is simply absent (not null). Hmm, in this case gross_pay is absent entirely.

Actually, maybe it means: when returning the answer for a clarify case, include the missing required value as explicit null (e.g., gross_pay: null) so the consumer knows what's missing. But the harness's allowed-fields list contradicts that.

The harness instruction: "Use exactly the required answer fields and only the clauses that directly govern the answer. Use the citation requirements in the supplied rules. Money must be strings with two decimal places. Allowed answer fields: calculate uses decision, currency, gross_pay, employee_contribution, allowed_deduction, taxable_pay, income_tax, net_pay, clauses; retrieve uses decision, field, value, clauses; clarify uses decision, reason_code, clauses."

"Allowed answer fields" — so for clarify, only decision, reason_code, clauses are allowed (plus optional explanation). I'll follow that. The explanation will name the blocker.

So final answer:

{
  "decision": "clarify",
  "reason_code": "MISSING_GROSS_PAY",
  "clauses": ["R1", "R7"],
  "explanation": "The payroll record for 2026-11 is missing gross_pay, which R1 requires. Please supply gross_pay so payroll can be calculated."
}

No monetary fields. No calculation needed. I don't need lookup_reference or calculate. The month is within 2026 so no NO_APPLICABLE_REFERENCE issue; the first check (missing gross_pay) triggers first anyway per the ordered checks.

Let me start by listing the documents to get the case inputs.

The payroll record has:
- payroll_month: 2026-11
- allowed_deduction: 0.00
- No gross_pay

Missing gross_pay → clarify with reason_code "MISSING_GROSS_PAY", citing R1 and R7.

Wait, but I should check the reference row for 2026-11? No — the order of checks in R7: first missing gross_pay (1), then missing payroll_month (2), then no reference row (3). Gross pay is missing, so the answer is clarify with MISSING_GROSS_PAY.

I don't need lookup_reference for this since the check order puts missing gross_pay first. But let me be careful — should I still verify the month has a reference row? The rules say check in this order: 1. Missing gross_pay first. So the answer is clarify MISSING_GROSS_PAY.

The fields for clarify: decision, reason_code, clauses, and optional explanation. Missing required values include explicit null — "Missing required values include explicit null." Hmm, for clarification cases, "For these cases do not return calculated monetary fields. Missing required values include explicit null." That suggests maybe include gross_pay: null? But the allowed answer fields are: calculate uses decision, currency, gross_pay, employee_contribution, allowed_deduction, taxable_pay, income_tax, net_pay, clauses; retrieve uses decision, field, value, clauses; clarify uses decision, reason_code, clauses. So clarify only uses decision, reason_code, clauses (plus optional explanation). The "explicit null" note might mean if I include gross_pay, it should be null — but the allowed fields for clarify don't include gross_pay. I'll stick with decision, reason_code, clauses, explanation.

Citations: R1 and R7.

The second-half reference covers 2026-07 through 2026-12, so 2026-11 is covered. But the first check in R7 is missing gross_pay. The record has no gross_pay field at all. So per R7 step 1: decision "clarify", reason_code "MISSING_GROSS_PAY", citing R1 and R7.

No calculation needed. Return the JSON object without Markdown fences or prose.

Let me start by listing the documents to get the case inputs.

The payroll record has:
- payroll_month: 2026-11
- allowed_deduction: 0.00
- No gross_pay!

According to R7, check order:
1. Missing `gross_pay`: return `decision: "clarify"` and `reason_code: "MISSING_GROSS_PAY"`, citing R1 and R7.

The record is missing gross_pay. So the answer is clarify with MISSING_GROSS_PAY.

The rules say: "Missing required values include explicit null." Hmm, "For these cases do not return calculated monetary fields. Missing required values include explicit null." That means in the clarification, missing values should be explicit null? The allowed answer fields for clarify: decision, reason_code, clauses. So no monetary fields. The "explicit null" refers to... perhaps if there were fields, they'd be null. But clarify only uses decision, reason_code, clauses. So I just return those.

Also, since the month is 2026-11, which is covered, I don't need to look up reference. Actually, the missing gross_pay check comes first anyway. But should I check whether the month has a reference row? The order: missing gross_pay is checked first (step 1), missing payroll_month step 2, no reference row step 3. Since gross_pay is missing, return MISSING_GROSS_PAY.

No need for lookup_reference or calculate. The answer:

{
  "decision": "clarify",
  "reason_code": "MISSING_GROSS_PAY",
  "clauses": ["R1", "R7"],
  "explanation": "..."
}

Include a short explanation naming the unresolved blocker and asking for its resolution.