| name | hr-regulated-domain |
| description | HR ERP regulated domains: wage/hour compliance, payroll calculation kernel, employment AI governance, MLOps/inference routing. Use for pay/time APIs, packages/payroll-calc, docs/compliance, lib/governance, screening/scoring. |
| risk | critical |
| minRiskTier | T3 |
| source | project |
| disable-model-invocation | true |
HR regulated domain
Collaboration plane: Do not invoke until phase 6 (revalidationConfirmed in plan/handoff). Use @hr-human-collaboration for phases 1–5.
Use this skill when
- Pay, time, earnings, jurisdiction matrices,
COMPLIANCE_* errors
packages/payroll-calc/, fingerprints, gross-to-net replay
- ML screening, scoring, governance APIs,
docs/ai-governance/
- Product MCP / inference paths with employment impact — co-load
@hr-product-mcp-governance and references/product-mcp.md when lib/copilot/** changes
Do not use this skill when
- Non-pay UI with no compliance or AI scoring surface
- T1/T2 work with no regulated path triggers
Instructions
- Load the relevant L3 reference(s) below — one domain at a time.
- Compliance + payroll kernel: read both when pay math executes.
- AI governance + MLOps: co-load when inference and scoring both change.
- Legal checklist →
specs/templates/legal-checklist.md; counsel confirms interpretation.
Resources
Limitations
- Engineering controls only — not legal advice
- Product runtime MCP still follows
docs/security/agent-mcp-threat-model.md