canonical: https://jentic.com/apis/fraudlabspro.com/fraudlabs-pro-fraud-detection

# FraudLabs Pro Fraud Detection

FraudLabs Pro Fraud Detection is the version 1 fraud-screening API for online merchants who need to detect chargeback risk on credit-card, PayPal, and cash-on-delivery transactions. The spec exposes two endpoints: `/v1/order/screen` scores a transaction in real time and returns a fraud verdict with the underlying signals, while `/v1/order/feedback` records the eventual outcome (approved, rejected, chargeback) so the per-merchant model improves with usage.

## For AI agents

Score a transaction for fraud risk and submit post-transaction feedback through FraudLabs Pro's v1 fraud-detection endpoints. Authentication is the API key on each request.

## Scope

Does not handle payment processing, refunds, or SMS verification - use for transaction fraud scoring and post-transaction feedback only.

## Capabilities

- Score a transaction for fraud risk via `/v1/order/screen` with order, billing, IP, and payment fields
- Receive a fraud score and recommended action (approve, review, reject) for each scored order
- Submit post-transaction feedback (approved, rejected, chargeback) via `/v1/order/feedback`
- Detect proxy, VPN, and high-risk IP signals as part of the screening response
- Improve the per-merchant fraud model over time by sending consistent feedback

## Use cases

### Real-Time Order Screening

An online merchant calls `/v1/order/screen` during checkout with the cart, billing and shipping addresses, customer IP, and payment method. FraudLabs Pro returns a numeric fraud score plus recommended action so the merchant can auto-fulfil low-risk orders, route medium-risk to manual review, and reject high-risk ones before settlement.

Example prompt: POST `/v1/order/screen` with the API key plus the order, billing, shipping, IP, and payment payload, then act on the recommended action in the response

### Post-Transaction Feedback Loop

Once the bank settles or charges back, the merchant calls `/v1/order/feedback` with the original order id and the realised outcome. FraudLabs Pro uses the feedback to tune the per-merchant model so future screens reflect what is actually fraud for that merchant's customer base.

Example prompt: POST `/v1/order/feedback` with the order id and the status (APPROVED, REJECTED, or CHARGEBACK) once the transaction settles

### Cash-On-Delivery Risk Filter

Cash-on-delivery merchants in Southeast Asia screen orders against FraudLabs Pro before dispatching the courier. The screen catches obvious fraud signals (mismatched IP and shipping country, proxy IPs, blacklisted phone numbers) so the merchant avoids the most expensive failure mode for COD: a delivered order the customer refuses.

Example prompt: POST `/v1/order/screen` with the COD order's phone, address, and IP and decline dispatch if the recommended action is REJECT

### AI Agent Fraud Triage

An agent triages a checkout queue, calling FraudLabs Pro through Jentic to score each order and submitting feedback once outcomes are known. The agent never sees the API key directly and surfaces a plain-language summary explaining the verdict for each order.

Example prompt: Search Jentic for 'score an order for fraud risk', load the `/v1/order/screen` schema, execute against each pending order, then post a summarised verdict to operations

## Key endpoints

| Method | Path | Description |
| --- | --- | --- |
| POST | `/v1/order/screen` | Score an order for fraud risk |
| POST | `/v1/order/feedback` | Submit post-transaction feedback (APPROVED, REJECTED, CHARGEBACK) |

## Key resources

- **Order Screening** — Score an order for fraud risk and retrieve recommended action and signals
- **Feedback** — Submit post-transaction outcomes to improve the per-merchant fraud model

## Why Jentic

- **Setup:** Wiring the FraudLabs Pro Fraud Detection API by hand means appending your license key to each request and building your own retry and error handling around the screen and feedback endpoints. Through Jentic you install once, import it from the API Directory, store the key once, and your agent calls it.
- **Permission scoping:** FraudLabs Pro Fraud Detection carries the order details in the request body rather than the URL path, so scoping is operations-only: you limit the agent to the operations it needs, such as scoring an order, and leave out posting feedback unless you add it. Every operation the agent can run is one you chose.
- **Credential handling:** Your FraudLabs Pro license key is stored once, encrypted, by your own Jentic One instance and injected at execution time. It never enters the agent's prompt, logs, or context.
- **Discovery method:** Agents search Jentic by intent such as 'score an order for fraud risk', and Jentic returns the matching FraudLabs Pro operation with its input schema so the agent works with the two endpoints directly rather than parsing the spec.

## Related APIs

- **FraudLabs Pro v2 API** — v2 of the same FraudLabs Pro product with refreshed paths and SMS verification endpoints
- **Sift** — Sift is a higher-end fraud and trust platform with richer event types and configurable ML
- **MaxMind minFraud** — MaxMind minFraud focuses on IP intelligence and device-based fraud signals

## FAQ

### What authentication does the FraudLabs Pro Fraud Detection API use?

The v1 fraud-detection API takes the FraudLabs Pro license key on each request as a parameter. Through Jentic, the key is stored encrypted in the vault and substituted at execution time so it never enters the agent's context.

### Can I score a credit-card transaction for fraud risk with this API?

Yes. POST to `/v1/order/screen` with the billing address, shipping address, customer IP, item list, and payment method. The response carries a numeric fraud score, a recommended action, and the underlying signals such as IP proxy detection and address mismatch.

### What are the rate limits for the FraudLabs Pro Fraud Detection API?

FraudLabs Pro applies monthly transaction quotas per plan rather than per-second rate limits. The free Micro plan covers a few hundred screens per month and paid plans scale into the millions; only `/v1/order/screen` calls count toward quota, not feedback.

### How do I send chargeback feedback through Jentic?

Search Jentic for 'send fraud feedback', load the `/v1/order/feedback` schema, and execute it with the order id from the original screen plus status set to CHARGEBACK. The feedback retrains the per-merchant model so future screens of similar customers are more accurate.

### Is the FraudLabs Pro v1 API free?

FraudLabs Pro offers a free Micro plan with a small monthly screen allowance, and paid plans scale by volume. The v1 fraud-detection endpoints share the same plan and quota as the rest of the FraudLabs Pro product.

### Should I use v1 or v2 of the FraudLabs Pro API?

New integrations should use v2 (under `/v2/order/screen`) for the latest signals and the SMS verification endpoints. The v1 endpoints under /v1 remain supported for legacy integrations and follow the same scoring contract.

### Can I limit what my agent is allowed to do with the FraudLabs Pro Fraud Detection API?

Yes. Because this API carries the order details in the request body rather than the URL path, scoping is operations-only, and your self-hosted Jentic One instance enforces the rules you set for which operations the agent may call. You can allow the agent to score an order via `/v1/order/screen` while leaving out `/v1/order/feedback`, or enable both if you want the agent to submit post-transaction outcomes. Every operation the agent can run is one you chose, and your license key stays with your own instance and is injected only at execution time.