Documentation

Requirements

• PHP 8.4+ with common extensions (mbstring, xml, curl, sqlite3/mysql)
• Composer 2.x
• Node.js 20+ & npm
• SQLite (development) or MySQL 8+ (production)
• Flux UI Pro v2 license — purchase at fluxui.dev

Quick Start

One-Click Store Setup

Choose an industry template and get products, collections, brands, URLs, and a matching storefront theme — all created instantly. Replace sample products with your real catalog and you're selling.

Automatic Processing

When you upload product images, WirePulse automatically:

• Resizes to max 1200×1500px (maintaining aspect ratio)
• Converts to WebP format (25-35% smaller than JPEG)
• Compresses at 85% quality (virtually invisible loss)

Upload JPEG, PNG, GIF, or WebP — the system converts everything to WebP automatically.

Creating a Product

Product Editor Tabs

Data Quality Score

The header badge shows how complete your product data is across 12 data points. Click to see missing fields.

Pricing

Each variant has three price fields:

How It Works

Setup

Add REMOVE_BG_API_KEY=your-key to .env. Free tier: 50 images/month. The AI prompt generator uses your existing AI provider.

Collections appear in storefront navigation and can be nested — e.g., "Electronics > Laptops > Gaming Laptops". Products can belong to multiple collections. Create them at Collections → Create Collection.

How It Works

Write a buying guide like "How to Choose the Right Tent" and link it to products in your catalog. Readers see the article content with live product cards showing current prices and stock status. Rich text editor, featured image upload (drag-and-drop), SEO fields with AI generation, and sortable product linking.

Blog Storefront

Product Linking

Both articles and blog posts can link to products. Linked products appear below the content as product cards with live prices and stock — bridging content marketing and commerce.

Stock is deducted when payment is confirmed, not when the order is created — abandoned checkouts don't lock inventory. Each order has a unique reference code (e.g., WP-A3K9M2X1). Click any order to see line items, addresses, payment info, and tax breakdown.

Creating a Draft Order

Multiple rules can stack — e.g., "10% off over 500 kr" and "Free shipping over 499 kr" both trigger on the same order. Configure at Settings → Discount Rules with optional date ranges and priority ordering.

How It Works

Stock is tracked per variant per location. The storefront shows the aggregate total. Configure at Settings → Locations. When an order is paid, stock is allocated from available locations — split across multiple fulfillments if needed.

Each supplier has contact details, currency, and configurable lead time. Suppliers are linked to product variants via a pivot table that stores supplier-specific SKUs, cost prices, minimum order quantities, and a "preferred supplier" flag.

Supplier-Variant Link

Creating a PO

Receiving Goods

Open the PO, enter received quantities per line, click "Receive Items". Stock is added to the destination location immediately. Supports partial deliveries — receive what arrived, leave the PO open for the rest.

Incoming Stock

When a PO is marked as "Ordered", the system sets an "incoming" count on each variant's inventory level at the destination. This is visible in the product editor — helping you see stock on hand vs. stock in transit, preventing double-ordering.

Transfer stock between any two locations. Select source, destination, product variant, and quantity. Ship deducts from source and sets "incoming" on destination. Receive adds stock at destination. Each transfer gets a unique reference (TRF-XXXXXXXX).

On Order Details, create fulfillments manually or let the system auto-allocate. Ship with optional tracking number/URL. Mark as delivered when the customer receives the package. Uses fulfillment.created and fulfillment.shipped webhook events.

Customer Flow

Customer opens a completed order in their portal → clicks "Request Return" → selects items, quantities, and reason → submits. Tracks status at /app/returns.

Admin Flow

Review at Sales → Returns → Approve/Reject → Mark as Received → Process Refund (gift card automatic or original method manual). Inventory is restocked on refund.

Webhook Events

The dashboard analyzes sales over the past 30 days and predicts when each variant will run out. Alerts appear in the amber panel if less than 30 days of stock remain.

How It Works

Click "Run Analysis" and the system calculates the capital value (stock x cost price) of every product, its sales velocity, days of stock remaining, 90-day ROI, and time since last sale. Products are classified as Healthy, Excess, Slow, or Dead — then displayed with a visual capital distribution bar showing where your money sits.

Product Classifications

Metrics Per Product

Each product shows: capital tied up, stock count, cost/retail price, margin %, units sold (30d/90d), daily velocity, days of stock remaining, 90-day ROI, and days since last sale. Sort by highest capital, lowest ROI, slowest velocity, or most days of stock to find your worst performers.

Smart Pricing Link

Every non-healthy product links directly to Smart Pricing for AI-recommended price adjustments. A 10-20% price reduction on dead stock is often enough to restart movement and recover capital.

Custom Segments

Build conditions with: Total Spent, Order Count, Last Order Days Ago, Avg Order Total, Registered Days Ago, Customer Group. Use Match ALL or Match ANY logic. Preview matches before saving.

How It Works

Create at Marketing → Product Feeds. Each feed gets a public URL at /feeds/{slug}.xml or .json. Submit the URL to the ad platform. Feeds regenerate automatically every 4 hours.

What Customers See

Admin Moderation

At Marketing → Reviews: 3 KPI cards (Pending, Approved, Rejected), status filter, and review cards with Approve/Reject/Delete actions. Each review shows stars, status, verified badge, full text, author info, and a direct link to the product.

SEO: Google Rich Snippets

Products with approved reviews automatically include AggregateRating JSON-LD structured data. Google can show star ratings directly in search results — significantly improving click-through rates.

How It Works

Commission Flow

Linking Codes to Discounts

One code can serve two purposes: PARTNER20 gives the customer 20% off (Discount) AND the affiliate 10% commission. Create the Discount first, then link it when creating the Affiliate. The customer sees a discount, you track the referral — everyone wins.

A nightly command (recommendations:generate) analyzes 90 days of orders. Products frequently bought together get a co-purchase score. On product pages, recommendations are shown in priority order:

Customers subscribe from a completed order → choose frequency → system auto-generates orders on schedule. B2B customers get instant processing with invoices. B2C customers receive a Stripe payment link.

Customer Self-Service

At /app/subscriptions: change frequency, update quantities, pause, resume, or cancel. Admins manage all subscriptions at Sales → Subscriptions.

Webhook Events

How It Works

Open any customer and the Timeline tab shows their complete story: a snapshot with CLV segment, total orders, spend, returns, and reviews — then a vertical timeline of every interaction in reverse chronological order. No more switching between Orders, Reviews, and Returns to piece together a customer's history.

Customer Snapshot

Timeline Events

• • Orders — Reference, status, total. Color-coded by status. Clickable link to order detail.
• • Reviews — Product name, star rating (★★★★☆), title or excerpt. Amber for 4-5★, red for 1-2★.
• • Returns — Order reference, return status, reason. Clickable link to return detail.
• • Abandoned Carts — Item count, cart value. Orange for abandoned, green if recovered.
• • Account Created — Registration date. Always the last event in the timeline.

Use Cases

• • Customer support — See full history before responding. "Last order 3 days ago, 2-star review yesterday, open return."
• • VIP management — Champions with 12 orders and 5-star reviews deserve personal attention.
• • Win-back decisions — At Risk with high-value history and no returns? Worth a campaign. Frequent returner? Maybe not.
• • Fraud patterns — 5 orders and 4 returns? The timeline makes it instantly visible.

Approved companies with payment terms and sufficient credit see "Pay on Account" at checkout. Order is placed immediately, invoice generated with due date. Stock decremented and fulfillments created. Overdue invoices flagged daily.

Enable in 4 Steps

Example Shopping Flow

Architecture

Powered by Laravel AI SDK with tool-calling. Five tools: SearchProducts, GetProductDetails, AddToCart, GetCartSummary, TrackOrder. Uses the cheapest model. Cart counter updates in real-time when AI adds items.

How It Works

Safety

Never suggests below cost price. Changes under 3% are filtered out. Runs weekly (Monday 04:00) or on-demand. Full audit trail of who applied which suggestion.

How It Works

Storefront

Available at /visual-search. Drag-and-drop image upload with preview, text search alternative, AI loading spinner, and product grid results. Purple-themed "Shop the Look" design.

How It Works

Click "Run Analysis" and the AI examines every customer's complete purchase history — total revenue, order count, average order value, purchase frequency, time between orders, and recency. It then classifies each customer into a segment, assesses churn risk, predicts future lifetime value, and recommends a specific action.

Customer Segments

Every customer with at least one order is classified into one of six segments:

Churn Risk Assessment

Independent of segment, each customer gets a churn risk level: Critical (very likely to leave), High (warning signs), Medium (stable but quiet), or Low (healthy, active). Filter by risk level to prioritize outreach.

AI Recommendations

Each customer profile includes a tailored action recommendation. Examples:

• • "Send 10% discount code to 47 At Risk customers"
• • "Invite to VIP program — 8 months of consistent high-value orders"
• • "Win-back campaign with 20% discount — no activity in 120 days"
• • "Product recommendation email — rising AOV suggests willingness to spend more"

Dashboard Overview

The dashboard shows four key metrics at the top: total analyzed customers, total predicted CLV (your future revenue potential), average CLV per customer, and at-risk revenue (how much you could lose without intervention). Below that, six clickable segment cards let you filter instantly. Each customer card shows predicted CLV, revenue, AOV, purchase frequency, recency, and the AI's analysis with recommended action.

How It Works

Click "Run Forecast" and the AI analyzes each product's sales history across 5 time windows (7d, 14d, 30d, 60d, 90d), day-of-week patterns, seasonal context, incoming stock from open purchase orders, and supplier lead times. It predicts demand for the next 14, 30, and 90 days — then calculates exactly when you'll run out and what to reorder.

Five Urgency Levels

What the AI Considers

• • Sales trends — 30d vs 90d velocity. Accelerating demand → higher forecast.
• • Day-of-week patterns — Weekend spikes factored into upcoming demand.
• • Seasonal events — Easter, Black Friday, Christmas detected within 30 days.
• • Incoming stock — Open purchase orders reduce stockout urgency.
• • Supplier lead time — Reorder deadline accounts for delivery delays.

AI Recommendations

Each product gets a specific, actionable recommendation:

• • "Order 200 units within 14 days to avoid stockout before Easter"
• • "Demand is accelerating +25%. Increase next order by 30%"
• • "90+ days of stock at current pace. 15% discount to move inventory"

Each card links directly to "Create Purchase Order" for one-click reordering.

Schedule

Runs automatically every Monday at 05:00, or on-demand via the dashboard. Works best with 90+ days of order history, preferred suppliers with lead times set, and active purchase orders for incoming stock tracking.

How It Works

Click "Find Bundles" and the AI scans 180 days of order data, finding product pairs that customers frequently buy together. It calculates the co-purchase rate (how often products appear in the same order), suggests a customer-friendly bundle name, recommends a discount percentage (5-20%), and explains why the bundle makes sense. Accept → the bundle is created as a draft. Publish → it appears on product pages and at checkout.

What You See

Workflow

• • Find Bundles → AI generates up to 20 suggestions ranked by co-purchase rate
• • Accept → Creates a draft bundle with products, prices, and discount calculated
• • Publish → Bundle goes live on product pages and checkout
• • Track → See how many times each bundle has been purchased
• • Dismiss → Reject suggestions that don't fit. Won't be suggested again

Where Bundles Appear

Published bundles show as "Frequently Bought Together" on the product pages of included products (with combined price and savings), and as upsell suggestions at checkout when the cart contains a bundle product. Both placements are configurable per bundle.

How It Works

Click "Analyze Orders" and the AI scans all orders from the last 30 days. For each order, it examines the items, quantities, customer return history, product categories, discount usage, and address patterns — then assigns a return probability (0-100%) and risk level with a specific proactive recommendation.

Risk Factors

Four Risk Levels

• • Critical (60%+) — Send personal email with sizing guidance
• • High (35-60%) — Include personalized note in package
• • Medium (15-35%) — Follow up 3 days after delivery
• • Low (<15%) — Standard fulfillment, no action needed

Self-Improving Accuracy

The system tracks actual outcomes — whether predicted high-risk orders were actually returned. Once 10+ outcomes are recorded, the dashboard shows prediction accuracy percentage. This feedback loop helps you understand how reliable the predictions are for your specific product mix and customer base.

How It Works

Create multiple storefronts in Settings → Storefronts. Each storefront gets its own domain, language, currency, customer price group, sales channel, and theme. When a visitor hits a storefront domain, the middleware automatically applies that storefront's configuration — locale, prices, theme, and product visibility. Inventory and the product catalog are shared across all storefronts.

Per-Storefront Configuration

Common Setups

• • B2B + B2C — Consumer store with retail prices, wholesale store with B2B price group. Same catalog, different pricing.
• • Multiple countries — store.de (German, EUR), store.no (Norwegian, NOK), store.com (English, USD).
• • Brand stores — Different theme and restricted catalog per brand.
• • Outlet — Separate storefront for clearance items with distinct look and restricted products.

Shared Resources

All storefronts share the same product catalog, inventory, customer database, and admin panel. A sale on one storefront reduces stock on all others. Orders are tagged with the storefront's sales channel for per-store revenue reporting. One admin team manages everything from a single backend.

Editor Layout

E-Commerce Modules

E-commerce modules pull live data from your catalog. Data is cached for 5 minutes.

How It Works

Blade Directive

@@ab('hero-cta')
  Buy Now!
@@else
  Shop Now
@@endab

Admin Dashboard

Side-by-side variant comparison with conversion rates, progress bars, and uplift calculation. Draft → Running → Completed lifecycle. One-click "Declare Winner" to end the test.

Charts

Revenue Over Time (line), Orders Over Time (bar), AOV (line), Customer Acquisition (dual-line with B2B highlight), Cart Recovery (stacked bar). All interactive with hover tooltips.

How It Works

Click "Run Audit" and the system scans every published product across six dimensions: images, descriptions, SEO, pricing, stock, and catalog completeness. You get an overall score (0-100) with a letter grade (A-F), category breakdowns with progress bars, a prioritized issues list, and AI-generated top 3 actions to improve your score.

Six Categories

Grading Scale

AI Actions

After scoring, the AI generates a 2-3 sentence summary of your store's state and the top 3 most impactful actions — ordered by impact. Example: "Add images to 14 products (Images score jumps from 60 to 90). Write meta titles for 10 products. Add descriptions to 3 products." Use these as your improvement to-do list.

Authentication

Create tokens at Settings → API Tokens. Include header: Authorization: Bearer YOUR_TOKEN. Rate limit: 60 req/min (120 req/min for admin tokens).

Endpoints

GET/POST   /api/v1/products       GET/POST  /api/v1/orders
GET/POST   /api/v1/customers      GET/POST  /api/v1/collections
GET/POST   /api/v1/brands         GET/POST  /api/v1/tags
GET/POST   /api/v1/discounts      GET       /api/v1/currencies
GET        /api/v1/channels       GET       /api/v1/dashboard/stats
GET        /api/v1/inventory

Filtering & Includes

GET /api/v1/products?filter[status]=published
GET /api/v1/products?include=variants,brand,media
GET /api/v1/orders?filter[status]=processing

Token Abilities (Scopes)

How It Works

Add a single script tag to any external page, then use the <wirepulse-buy> web component to embed a product card or buy button. Customer clicks "Buy Now" → Stripe Checkout opens in a new tab → payment completes → order appears in your admin panel with the referrer URL.

Two Layouts

Widget Options

Technical

Pure vanilla JavaScript — no dependencies. Renders inside Shadow DOM for complete CSS isolation. CORS-enabled public API with 60 req/min rate limiting. Orders tagged with source: embed and the referrer URL for attribution tracking. Works on any website: WordPress, Webflow, Squarespace, static HTML, or any CMS.

Setup

Settings → Webhooks → "Add Webhook" → enter endpoint URL → select events → Test → Save. Retries: 3 attempts with backoff (10s, 1m, 5m).

Available Events (21)

Signature Verification

Each payload is signed with HMAC-SHA256. Verify the X-WirePulse-Signature header against your webhook secret.

$payload = file_get_contents('php://input');
$expected = hash_hmac('sha256', $payload, $secret);
if (hash_equals($expected, $signature)) {
    // Webhook is authentic
}

Plugin Structure

plugins/my-plugin/
├── manifest.json                    # Plugin metadata
└── src/
    └── MyPluginServiceProvider.php  # Registration logic

Extension Points (Hooks)

Example Registration

$manager = app(PluginManager::class);
$manager->registerHook('admin.sidebar.items', fn () => [
    'label' => 'My Feature',
    'icon'  => 'star',
    'route' => '/wirepulse/my-feature',
]);

Environment Variables

# Your license key from wirepulse.org
WIREPULSE_LICENSE_KEY=

# License server URL (default: https://wirepulse.org)
WIREPULSE_LICENSE_SERVER=https://wirepulse.org

# Cache TTL in hours (default: 24)
WIREPULSE_LICENSE_CACHE_TTL=24

Daily Verification

A scheduled command (license:verify) runs daily and contacts the license server at POST /api/v1/verify. The server validates the key and returns the tier (free/pro/enterprise) and expiry date. Results are cached locally.

Manual Verification

php artisan license:verify

# Output (valid):
# License valid — Tier: pro, Expires: 2027-03-30

# Output (no key):
# License invalid or not configured — running in free/community mode.

Graceful Degradation

Admin Banner

A small banner at the top of the admin panel indicates your license status:

Managing Your Subscription

Log in at wirepulse.org/app to view your license key, manage your subscription via Stripe (upgrade, downgrade, cancel), and download invoices. Tier changes take effect on the next license verification.