Can I vibe code NoFraud?
nofraud.com · fraud · usage-based · rev-share
The verdict
NOT REALLY — THE UI ISN'T THE MOAT
NoFraud (now Wyllo) is two distinct things sold in one contract: a risk-scoring platform and an insurance policy. The software layer runs pre-capture or post-checkout risk models using IP geolocation, device fingerprinting, BIN lookups, and behavioral tracking to approve or block orders. The business layer charges a 1.0% to 1.5% take rate (with a $250-$450/month minimum for stores doing under $50k/mo) in exchange for indemnifying the merchant against fraudulent chargebacks on approved orders. You can clone the software layer in two weeks using APIs like MaxMind, FingerprintJS, or Sift. You cannot clone the insurance policy. No software builder can underwrite credit card chargeback liability out of the box. Merchants do not pay NoFraud primarily for the Python risk engine; they pay for the financial balance sheet that absorbs chargebacks and the ops team that manually reviews edge-case orders. Build the code if you want a custom risk filter, but buy the software if you want chargeback indemnity.
- Replaces
- $250/mo
- MVP build time
- 2-3 weeks
- Full replacement
- 12-18 months
- Verdict
- NOT REALLY
What it really costs
| Screening Only (ProtectionLite) | free / quote | Up to 100 screened orders/month |
| w/ $500 Guarantee | $250/mo | 1% of revenue ($250/mo minimum) for <$50k/mo revenue |
| w/ $1000 Guarantee | $350/mo | 1.25% of revenue ($350/mo minimum) for <$50k/mo revenue |
| w/ $2000 Guarantee | $450/mo | 1.50% of revenue ($450/mo minimum) for <$50k/mo revenue |
Screening-only plan is free for up to 100 orders/mo. Chargeback guarantee plans cost 1.0%-1.5% of revenue with monthly minimums starting at $250/mo for stores under $50k/mo.
- Captured
- 2026-08-04 (3 days ago)
- Verified by
- human
- Source
- apps.shopify.com
Assumptions: Screening-only plan is free for up to 100 orders/mo. Chargeback guarantee plans cost 1.0%-1.5% of revenue with monthly minimums starting at $250/mo for stores under $50k/mo.
The one-shot build prompt
Build a Node.js/TypeScript fraud evaluation microservice for Shopify and custom e-commerce stores with webhook and admin dashboard integrations.
DATA MODEL:
1. Store: id, shopify_domain, api_key, risk_threshold_review (0-100), risk_threshold_cancel (0-100), auto_cancel_enabled (boolean).
2. OrderEvaluation: id, store_id, external_order_id, risk_score (0-100), decision (PASS, REVIEW, FAIL), indicators (JSON array of flags), raw_signals (JSON), status (PENDING, DECIDED, REVIEWED), human_override (boolean), created_at.
3. RiskRule: id, store_id, field (billing_shipping_distance, proxy_detected, email_domain_age, velocity_count, order_value_usd), operator, threshold, score_impact.
CORE WORKFLOW:
1. Webhook Ingestion: Listen for shopify/orders/create. Extract IP address, user_agent, billing_address, shipping_address, customer_email, total_price, credit_card_bin.
2. Signal Gathering Pipeline:
- Call MaxMind minFraud API (or mock wrapper) with order IP to detect VPN/Proxy/Tor usage and IP-to-billing distance.
- Calculate Haversine distance between billing postal code coordinates and shipping postal code coordinates.
- Run email reputation check (e.g., disposable email check, domain registration age).
- Compute velocity check: query Redis for count of orders from the same IP or customer email in the last 1 hour.
3. Decision Engine:
- Calculate composite risk score (0-100) based on weighted sum of signal impacts and active RiskRules.
- If score >= risk_threshold_cancel and auto_cancel_enabled, send cancellation request via Shopify Admin API (orders/{id}/cancel) and add order tag 'NoFraud_Failed'.
- If score >= risk_threshold_review, flag order for review, set decision = REVIEW, and tag order 'NoFraud_Review'.
- Otherwise set decision = PASS and tag order 'NoFraud_Passed'.
4. Operational Dashboard (React/Next.js):
- Orders review queue displaying all 'REVIEW' status orders with map overlay comparing IP location, billing address, and shipping address.
- Real-time action buttons: 'Approve & Pass' (updates status, removes review tag) and 'Reject & Cancel' (triggers Shopify cancellation API).
OUT OF SCOPE:
- Chargeback financial guarantee payouts or insurance logic.
- Credit card processor representment filing workflows.
- Browser fingerprinting JS payload generation (rely on raw request header metadata and API vendor scores).
FAILURE MODES TO HANDLE:
- Shopify webhook duplicate deliveries: use order ID idempotency keys in Redis before evaluation.
- API provider timeout (e.g., MaxMind down): fallback to local rule-based evaluation (distance + velocity) without failing open.$ each button prefixes agent-specific run instructions · build your own product, never copy proprietary code, trademarks or designs
Scorecard
What you can actually replace
- ✓Automated order risk scoring based on standard rules (IP, distance, proxy, email).
- ✓Shopify order tagging and automatic cancellation API triggers.
- ✓Manual order review dashboard with signal visualizations.
- ✓Basic velocity and heuristic-based risk engine.
What you lose
- ×100% financial reimbursement guarantee on fraudulent chargebacks passed by the engine.
- ×24/7 human analyst team conducting manual order reviews on borderline transactions.
- ×Network-level risk detection trained on shared cross-merchant fraud signals.
- ×Direct chargeback dispute handling and representment operations.
Why people still pay — the real moats
Moats
- — Balance sheet capital to guarantee chargeback reimbursements at scale.
- — Operational infrastructure of human fraud analysts performing 24/7 manual reviews.
- — Consolidated network intelligence across thousands of high-volume e-commerce storefronts.
Hard parts
- — Real-time evaluation latency under sub-500ms constraints for synchronous checkout gating.
- — Cross-merchant device fingerprint resolution and identity mapping across disjointed stores.
- — Balancing false-positive friction against chargeback risk without access to issuing bank internal signals.
- — Continuous manual review queue staffing for ambiguous high-ticket orders.
- — Financial risk management and loss reserves required to pay out chargeback guarantee claims.
- — Filing and managing reimbursement claims with acquiring banks and merchant processors.
Network effects you cannot generate
- — Cross-merchant identity graph (detecting a fraudster who struck Store A when they attempt to buy at Store B).
- — Aggregated issuer/bin risk profile data derived from millions of monthly transactions across the merchant portfolio.
Build this instead
Custom Pre-Capture Fraud Rules Engine
Build an edge worker that combines MaxMind MinFraud, Sift, or Radar with Shopify Checkout Extensions to inject pass/review/fail signals into custom payment rules before capture.
Automated Chargeback Evidence Generator
Build an automated evidence bundler for shopify/stripe chargebacks that combines shopify order data, carrier delivery signature proofs, and IP logs into a dispute PDF.
Consolidated Merchant Risk Consortium Engine
Build a cross-store device and email reputation graph using privacy-preserving bloom filters or hashed identifiers across an agency's merchant portfolio.
Prior art — do not start from zero
Open source alternatives to NoFraud
Self-hostable projects that cover most of the same ground. Free licence, your infrastructure, your on-call.
Have you actually replaced it?
Related products in this category
Signifyd is a fraud-decision engine paired with a chargeback insurance policy that guarantees approved orders against fraud loss.
$1,500/mo
Trustfull aggregates digital footprint signals—phone, email, IP, domain, and device intelligence—to calculate real-time fraud risk scores for digital onboarding and transactions.
$500/mo
Enterprise e-commerce fraud decision engine that combines real-time device fingerprinting, transaction scoring, and chargeback financial guarantees.
$10,000/mo
FAQ
+Can I really replace NoFraud with an AI-generated app?
NOT REALLY — YOU CAN BUILD THE SCORING ENGINE, BUT YOU CANNOT CODE AN INSURANCE BALANCE SHEET. NoFraud (Wyllo) is fundamentally an insurance product wrapped in an API. While the order scoring and Shopify tagging features can be cloned with standard LLM tools in a couple of weeks, you cannot write code that underwrites thousands of dollars in credit card chargebacks. An MVP takes roughly 2-3 weeks; matching the product properly is closer to 12-18 months.
+How long does it take to rebuild NoFraud?
A usable internal version: 2-3 weeks. A version you would sell or bet a business on: 12-18 months, mostly spent on real-time evaluation latency under sub-500ms constraints for synchronous checkout gating..
+What do you actually lose by leaving NoFraud?
100% financial reimbursement guarantee on fraudulent chargebacks passed by the engine. 24/7 human analyst team conducting manual order reviews on borderline transactions. Network-level risk detection trained on shared cross-merchant fraud signals.
+Is it legal to build a NoFraud alternative?
Building a competing product with your own code is normal competition. Copying their code, trademarks, brand assets or scraping their platform is not. Use the prompt to build your own implementation of common features.
Written by Andrea Saccà — 18 years in the Magento ecosystem. Last reviewed 2026-08-04.
Scores are computed, not typed. Read the methodology.
One e-commerce SaaS teardown every week.
Honest verdicts, build prompts and overlooked vertical SaaS opportunities. No tracking pixels, no drip sequence, unsubscribe in one click.
free forever · no third-party tracking · the prompts stay public