Back to Home
Johnza06 icon

advance-fraud-analyst

Verified Safe

by Johnza06

Overview

The application provides a web-based interface for fraud detection and risk analysis across multiple data types (transactions, KYC, sanctions, credit risk) using rule-based engines and an AI consultant agent powered by Large Language Models.

Installation

Run Command
python app.py

Environment Variables

  • FIREWORKS_API_KEY
  • HF_TOKEN
  • ENABLE_MCP
  • MCP_SANCTIONS_URL
  • MCP_HIGH_RISK_MCC_URL
  • MCP_AUTH_HEADER
  • TTP_BLOCK_LEVEL

Security Notes

The application handles secrets (API keys) via environment variables, which is good practice. It includes a TTP (Tactics, Techniques, and Procedures) Guard to detect prompt injections and other adversarial inputs, enhancing security. External calls to LLM providers and optional MCP endpoints are made over HTTPS with timeouts. No 'eval' or direct code injection vulnerabilities were identified. Potential minor risk if MCP URLs are misconfigured to untrusted endpoints, but these are controlled by environment variables.

Similar Servers

Stats

Interest Score0
Security Score8
Cost ClassMedium
Avg Tokens200
Stars0
Forks0
Last Update2026-01-19

Tags

Fraud DetectionRisk ManagementAI/MLData AnalysisLLM Agent