Let me start by listing the documents to see what's in this case.

The payroll record has only `payroll_month: "2026-02"` and no `gross_pay`. Per R7 order:

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

The payroll month is present (2026-02, valid format and covered by reference rows), so the first check that fails is missing gross_pay.

The answer fields for clarify: decision, reason_code, clauses. Optional explanation string. No monetary fields.

Clauses: R1 and R7.

The record has payroll_month "2026-02" but no gross_pay. Per R7 check order:

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

The month is present and covered (2026-02 falls in the first-half row's months 2026-01 through 2026-06), so check 1 applies: missing gross_pay comes first in the order.

Return clarify JSON with decision, reason_code, clauses. No monetary fields. Include a short explanation naming the unresolved blocker.

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

The payroll record has only `payroll_month: "2026-02"` and no `gross_pay`. Per R7 step 1: Missing `gross_pay` → clarify with reason_code "MISSING_GROSS_PAY", citing R1 and R7.

Wait, order of checks: R7 says check in order:
1. Missing gross_pay: clarify MISSING_GROSS_PAY
2. Missing payroll_month: clarify MISSING_PAYROLL_MONTH
3. Month without reference row: NO_APPLICABLE_REFERENCE
4. Conflicting gross_pay

Here gross_pay is missing, so reason_code MISSING_GROSS_PAY. The payroll month is present and within 2026, but the first check triggers.

The answer format for clarify: decision, reason_code, clauses, optional explanation. "Missing required values include explicit null." Hmm — "For these cases do not return calculated monetary fields. Missing required values include explicit null." So maybe include gross_pay: null? The allowed answer fields for clarify are decision, reason_code, clauses (plus optional explanation). The instruction says "Missing required values include explicit null" — this suggests if a field like gross_pay is missing, include it as explicit null. But the allowed fields list doesn't include gross_pay for clarify. Hmm, "Do not include input-only fields such as payroll_month, contract_gross_pay, or amendment." gross_pay is an answer field for calculate. For clarify, allowed fields are decision, reason_code, clauses. I think including gross_pay: null might be acceptable given "Missing required values include explicit null". But the safest interpretation: the allowed answer fields for clarify are decision, reason_code, clauses, and optional explanation. The "explicit null" instruction likely refers to not substituting zero — i.e., don't guess. Since the schema for clarify only allows decision, reason_code, clauses (and explanation), I'll stick to those. Actually, let me reconsider: "Missing required values include explicit null. Do not guess a value or substitute zero." This is guidance on how to handle missing values in the JSON — include them as null rather than omitting or guessing. But "Use exactly the required answer fields" for clarify is decision, reason_code, clauses. Adding gross_pay: null could violate "Use exactly the required answer fields". The phrase "and only the clauses that directly govern the answer" refers to clauses.

Hmm, there's tension. I think the safest is to follow the allowed fields strictly: decision, reason_code, clauses, explanation. The "explicit null" line likely applies to records where a value is missing but the schema expects it — e.g., in calculate answers if something were missing. But for clarify, the fields are fixed. I'll include explanation naming the blocker and asking for resolution.

Do I need to check the reference row? No — the blocker is missing gross_pay, which is check #1. No calculation needed. Cite R1 and R7.

The output must be one JSON object without Markdown fences or surrounding prose.

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

The payroll record has only `payroll_month: "2026-02"` and no `gross_pay`. Per R7 check order:

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

The month 2026-02 is covered (2026 is covered), but the check order says missing gross_pay is checked first. So the answer is clarify with MISSING_GROSS_PAY.

The clarification should include a short explanation naming the unresolved blocker and asking for its resolution. Fields: decision, reason_code, clauses. Missing required values include explicit null — but for clarify, the fields allowed are decision, reason_code, clauses (and optional explanation). The rule says "Missing required values include explicit null" — but the allowed answer fields for clarify don't include gross_pay. Actually the allowed fields list says: "clarify uses decision, reason_code, clauses." So I should not include gross_pay field. The "explicit null" mention might apply to calculate cases with missing values, but since clarify only allows those fields, I'll stick to decision, reason_code, clauses, and explanation.

Do I need lookup_reference? The month is 2026-02, which is covered by reference rows. But since gross_pay is missing, check #1 triggers first — no need to look up reference for the clarify decision. The check order is: missing gross_pay first. So no lookup needed.

Citations: R1 and R7.

Return one JSON object without Markdown fences or surrounding prose.