# BRAINIAC Refund Abuse Pattern Detector

> BRAINIAC Refund Abuse Pattern Detector is a paid API for AI agents from www.brainiac.software, paid per call via x402, $0.04/call, status unknown (last checked 2026-11-12).

Analyzes a set of orders to detect refund abuse and repeat-refund fraud patterns for a customer or batch of transactions.

## Facts

- Endpoint: POST https://www.brainiac.software/api/v1/ecommerce-fraud/refund-abuse-pattern-detect?utm_source=zero.xyz
- Price: $0.04/call
- Payment: x402
- Status: unknown
- Last checked: 2026-11-12
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/brainiac-refund-abuse-pattern-detector-c5a7192b
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_pOFda0gHJXdVOowdGnDEK

Status and success rate cover calls made through Zero and Zero's own probes. Third-party monitors may report differently.

## How to call it through Zero

Zero handles the 402 payment challenge and records the run. With the Zero CLI installed (`npm i -g @zeroxyz/cli`):

```sh
zero fetch --capability brainiac-refund-abuse-pattern-detector-c5a7192b -d '<json body>'
```

Example prompt: Check these 15 orders for refund abuse patterns — the customer has had 5 refunds out of 8 orders in the last 30 days. Flag any repeat refund or high refund rate signals and tell me whether this looks like policy abuse.

## When to prefer this

Use this endpoint when you need a lightweight, no-account, pay-per-call refund fraud signal on a small batch of orders (up to 200) without integrating a full fraud platform. Ideal for AI agents that need to flag suspicious refund patterns in real time during a support or review workflow, or for periodic batch audits of exported order data. Prefer this over manual review or heavyweight fraud suites when speed, low cost ($0.04/call), and zero onboarding friction matter. Not suited for real-time transaction blocking or when you need historical data beyond what you supply.

## Known failure modes

- Fewer than the minimum required orders supplied — signals cannot be computed
- Invalid order status values or missing required fields in order objects
- CSV or JSON payload exceeds 200,000 character limit
- Orders array exceeds 200-item limit
- Rules parameters outside allowed min/max bounds — validation error returned
- Ambiguous or missing customer identifiers prevent grouping — patterns may be incomplete
- Payment failure via x402 if USDC balance is insufficient

## How this service works

Pay-per-call APIs for AI agents: OFAC sanctions screening, compliance checks, token security, and Base chain data via x402. No account. Fixed USDC price. MCP-ready.

## Output

Returns a JSON object with a risk level (e.g. 'review'), a status, an array of signals (each with a code, rule, detail, affected order IDs, and a confirmed flag), an array of findings with kind and subject, a checkedAt timestamp, boolean flags (fraudConfirmed, orderBlocked, accountAccessed, paymentProcessed), and a limitations list clarifying that only supplied data is analyzed and no blocking action is taken.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "csv": {
   "type": "string",
   "maxLength": 200000
  },
  "file": {
   "type": "object",
   "required": [
    "mediaType",
    "fileBase64"
   ],
   "properties": {
    "filename": {
     "type": "string",
     "maxLength": 180,
     "minLength": 1
    },
    "mediaType": {
     "enum": [
      "text/plain",
      "text/csv",
      "application/json"
     ],
     "type": "string"
    },
    "fileBase64": {
     "type": "string",
     "maxLength": 12000000,
     "minLength": 1
    }
   },
   "additionalProperties": false
  },
  "json": {
   "type": "string",
   "maxLength": 200000
  },
  "text": {
   "type": "string",
   "maxLength": 200000
  },
  "rules": {
   "type": "object",
   "properties": {
    "orders": {
     "type": "integer",
     "maximum": 50,
     "minimum": 3
    },
    "quantity": {
     "type": "integer",
     "maximum": 1000000,
     "minimum": 1
    },
    "smallAmount": {
     "type": "number",
     "maximum": 1000000,
     "minimum": 0
    }
   },
   "additionalProperties": false
  },
  "orders": {
   "type": "array",
   "items": {
    "type": "object",
    "properties": {
     "id": {
      "type": "string",
      "maxLength": 80,
      "minLength": 1
     },
     "ip": {
      "type": "string",
      "maxLength": 80
     },
     "card": {
      "type": "string",
      "maxLength": 40
     },
     "promo": {
      "type": "string",
      "maxLength": 40
     },
     "amount": {
      "type": "number",
      "maximum": 1000000000,
      "minimum": 0
     },
     "status": {
      "type": "string",
      "maxLength": 40
     },
     "address": {
      "type": "string",
      "maxLength": 200
     },
     "country": {
      "type": "string",
      "maxLength": 40
     },
     "created": {
      "type": "string",
      "maxLength": 40
     },
     "currency": {
      "type": "string",
      "maxLength": 12
     },
     "customer": {
      "type": "string",
      "maxLength": 160
     },
     "quantity": {
      "type": "integer",
      "maximum": 1000000,
      "minimum": 0
     },
     "billCountry": {
      "type": "string",
      "maxLength": 40
     }
    },
    "additionalProperties": false
   },
   "maxItems": 200
  },
  "window": {
   "type": "integer",
   "maximum": 10080,
   "minimum": 1
  }
 }
}
```

## Response schema (JSON Schema)

```json
{
 "type": "json",
 "example": {
  "level": "review",
  "status": "checked",
  "signals": [
   {
    "code": "repeat_refund",
    "rule": "min_refunded_orders",
    "detail": "Customer C-1 has 3 supplied orders with status refunded.",
    "orders": [
     "R-1",
     "R-2",
     "R-3"
    ],
    "confirmed": false
   },
   {
    "code": "refund_rate",
    "rule": "refunded_share_at_least_half",
    "detail": "Customer C-1 has 3 refunded orders out of 4 supplied orders. The share is at least half and at least 3 refunds.",
    "orders": [
     "R-1",
     "R-2",
     "R-3"
    ],
    "confirmed": false
   }
  ],
  "findings": [
   {
    "code": "repeat_refund",
    "kind": "indicator",
    "detail": "Customer C-1 has 3 supplied orders with status refunded.",
    "subject": "R-1"
   },
   {
    "code": "refund_rate",
    "kind": "indicator",
    "detail": "Customer C-1 has 3 refunded orders out of 4 supplied orders. The share is at least half and at least 3 refunds.",
    "subject": "R-1"
   }
  ],
  "checkedAt": "2026-11-04T09:00:00.000Z",
  "limitations": [
   "Patterns are calculated only from the orders supplied in the request.",
   "One signal is an indicator. It does not mean the customer is fraudulent.",
   "No payment is processed, no order is blocked, and no merchant account is accessed.",
   "A full card number is not copied. A statistical outlier is not confirmed fraud."
  ],
  "orderBlocked": false,
  "fraudConfirmed": false,
  "accountAccessed": false,
  "paymentProcessed": false
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/brainiac-refund-abuse-pattern-detector-c5a7192b/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from www.brainiac.software](https://www.zero.xyz/host/www.brainiac.software/llms.txt)
