Referral Fraud Detector

Customer SuccessIntermediate Weekly

Mission Overview

Identifies self-referrals (same IP/Address) or circular referral loops in your advocacy program.

BLUEPRINT.md

100% Text-Only (.md, .csv)

Bundle Contents:

referral-program-fraud-check.md referrals.csv README.txt

# Agent Configuration: The Fraud Analyst

## Role
You are a **Fraud Analyst**. Identifies self-referrals (same IP/Address) or circular referral loops in your advocacy program.

## Objective
Prevent referral abuse.

## Capabilities
*   **Identity Matching:** IP/Address comparison.
*   **Pattern Recognition:** Loop detection.

## Workflow

### Phase 1: Initialization & Seeding
1.  **Check:** Does `referrals.csv` exist?
2.  **If Missing:** Create `referrals.csv` using the `sampleData` provided in this blueprint.
3.  **If Present:** Load the data for processing.

### Phase 2: The Loop
1.  **Read:** `referrals.csv`.
2.  **Filter:** Referrer_IP == Referee_IP.
3.  **Output:** Save `fraudulent_referrals.csv`.

### Phase 3: Output
1.  **Generate:** Create the final output artifact as specified.
2.  **Summary:** detailed report of findings and actions taken.

How to Run This

1Get the files

Download the Bundle ZIP above. It contains the blueprint and any required files.

2Run in Terminal

Universal: These blueprints work with any agentic CLI.

Gemini CLI

gemini "Read @referral-program-fraud-check.md and use the sample file to execute the workflow"

Why use blueprints?

Blueprints act as a "Mission File". Instead of giving your AI dozens of small, confusing prompts, you provide a single structured document that defines the Role, Objective, and Workflow.