Prior Authorization Review Skill

Automate payer review of prior authorization requests using AI-powered clinical reasoning and coverage policy matching to aid the payer's determination and workflow.

Overview

This Claude Code skill helps automate the health insurance payer professional's prior authorization (PA) review process. It processes incoming PA requests from healthcare providers, validates medical necessity against coverage policies, and generates proposed authorization decisions with complete documentation for the professional's review and sign-off.

Target Users: Health insurance payer organizations (Medicare Advantage, Commercial, Medicaid MCOs)

Value Proposition:

Consistent, auditable decision-making
Structured documentation for review processes
Free clinical reviewers to focus on complex cases

Important Disclaimers

DRAFT RECOMMENDATIONS ONLY: This skill generates draft recommendations only. The payer organization remains fully responsible for all final authorization decisions.

HUMAN REVIEW REQUIRED: All AI-generated recommendations require review and confirmation by appropriate professionals before becoming final decisions. Users may accept, reject, or override any recommendation with documented justification.

AI DECISION BEHAVIOR: In default mode, AI recommends APPROVE or PEND only - never recommends DENY. Decision logic is configurable in the skill's rubric.md file.

COVERAGE POLICY LIMITATIONS: Coverage policies are sourced from Medicare LCDs/NCDs via CMS Coverage MCP Connector. If this review is for a commercial or Medicare Advantage plan, payer-specific policies may differ and were not applied.

Subskill 1: Intake & Assessment

Purpose: Collect prior authorization request information, validate credentials and codes, extract clinical data, identify applicable coverage policies, assess medical necessity, and generate approval recommendation.

What it does:

Collects PA request details (member, service, provider, clinical documentation)
Validates provider credentials via NPI MCP Connector
Validates ICD-10 codes via ICD-10 MCP Connector (batch validation)
Validates CPT/HCPCS codes via WebFetch to CMS Fee Schedule
Searches coverage policies via CMS Coverage MCP Connector
Extracts structured clinical data from documentation
Maps clinical evidence to policy criteria
Performs medical necessity assessment
Generates recommendation (APPROVE/PEND)

Output: waypoints/assessment.json - Consolidated assessment with recommendation

Human Decision Point: After assessment is generated, the professional reviews the AI recommendation and supporting evidence before proceeding to the decision phase.

Subskill 2: Decision & Notification

Purpose: Finalize authorization decision with human oversight and generate provider notification.

What it does:

Loads assessment from Subskill 1
Presents recommendation for human review
Human actively confirms, rejects, or overrides the recommendation
Generates authorization number (if approved) or documentation requests (if pended)
Creates provider notification letter
Documents complete audit trail

Output:

waypoints/decision.json - Final decision record
outputs/notification_letter.txt - Provider notification

Human Decision Point: The professional must explicitly confirm, reject, or override the AI recommendation. No decision is finalized without human action.

Installation

See SKILL.md for complete installation and usage instructions.

Sample Data

Sample case files are included in assets/sample/ for demonstration purposes. When using sample files, the skill operates in demo mode which:

Skips NPI MCP lookup for the sample provider only
Executes all other MCP calls (ICD-10, CMS Coverage) normally
Demonstrates the complete workflow with a pre-configured lung biopsy case

Prior Authorization Review