What problem does it solve?
This Skill helps assess insurance claim fraud risk by analyzing the consistency of claim materials, abnormal behavior patterns, and historical records to produce a risk score and a list of suspicious signals for investigation support.
Core Features & Use Cases
- Multi-dimensional consistency checks: Detect inconsistencies in dates, parties, amounts, hospitals, diagnoses, and signatures/stamps.
- Behavior pattern analysis: Identify high-risk patterns such as short-term post-purchase incidents and high-frequency claims.
- Investigation-ready risk output: Provide a standardized 0–100 fraud risk score, risk level (low/medium/high/critical), evidence summary, and targeted investigation suggestions.
- Sidecar monitoring mode: Recompute or incrementally update risk as new data arrives across the claim lifecycle.
Use cases:
- Fraud screening for claim submissions (potential fake claims or骗保).
- Decision support for suspicious-claim investigations and compliance management.
- Triggering alerts when risk crosses predefined thresholds.
Quick Start
Ask the skill to evaluate the fraud risk for the submitted claim case materials and output a 0–100 score, risk level, suspicious signals, and investigation recommendations.
