/comp-benchmark

Purpose

/comp-benchmark produces a public-source compensation reference grid for a role by mapping the role to a Standard Occupational Classification (SOC) code, querying whitelisted public wage sources (BLS OEWS, BLS QCEW, OPM, O*NET, US Census ACS, state/city pay-transparency disclosures, municipal pay schedules), applying documented geography and seniority adjustments, and checking jurisdiction pay-transparency obligations. It produces a market reference grid + methodology worksheet, not a recommended offer. It is a drafting and triage aid, not HR advice, not employment-law advice, not pay equity analysis.

What it IS: a structured first pass that produces a market reference grid (10th / 25th / 50th / 75th / 90th percentile base pay bands) a human compensation analyst or HR leader can review, interrogate, and own, with every source, SOC code, adjustment factor, and data quality flag surfaced explicitly.

What it is NOT: a pay equity audit (see hard-scoped-out section below and future pay-equity-audit), a recommended offer, an individual offer construction tool, an internal equity analysis, a retention risk assessment, a variable-comp modeler (bonus / equity / commission), or a substitute for employment counsel on jurisdiction-specific pay-transparency obligations.

This skill is governed by hr-ai-governance pack v1.0.1. Every benchmark run inherits the proxy register (Section 3.2), audit log schema (Section 4.1), jurisdiction matrix (Section 8.1), and review cadence (Section 11.1 subsequent-similar 72h) from the pack. See the Pack Inheritance section below for the exact dependency map.

When to Use

Invoke /comp-benchmark when you need to: