AI assistance
Optional. Connect your own AI provider to get insight reports and drafts of wishlist copy and emails. Wishlists, sharing, alerts and reports never depend on it.
Turning it on
- AI Integration tab
In AI Features, turn on Enable AI features (while it's off, no AI code runs and no request is ever made). In AI Provider, pick a provider card; in Connection, paste your key and click Save Key, choose a model, Save Changes, then Test Connection ("Connected to …").
- AI Studio tab
Generate reports, ask questions and draft copy and emails.
The AI Integration tab's cards, in order: Wishlist Activity, AI Features, AI Provider, Connection, Requests & Limits. Its description: "Optional. Connect your own AI provider account and choose what it powers. You pay your provider directly, and your API key stays encrypted on this server."
- Provider cards with their status
- API key, model and Test Connection
Providers
| Provider | Default model |
|---|---|
| OpenAI | gpt-4.1-mini |
| Anthropic | claude-haiku-4-5-20251001 |
| Google Gemini | gemini-2.5-flash |
| Groq | llama-3.3-70b-versatile |
| OpenRouter | openai/gpt-4.1-mini |
| Mistral | mistral-small-latest |
| Custom endpoint | Any OpenAI-compatible endpoint. Enter its HTTPS Endpoint base URL (e.g. https://gateway.example.com/v1) and the model ID; plain HTTP is refused |
| WordPress AI | The AI connectors set up in WordPress itself (WordPress 7.0+); no key stored. The card appears only when WordPress provides them |
Each provider card has a "How to connect" guide with links to get a key. Each provider keeps its own key: Remove Key deletes a saved one. Model lists the models your key can use (Refresh models reloads them), or choose Other model ID... to type one; leave it on "Provider default" to use the model above.
You pay your provider directly. Keys are encrypted on your server, never shown again and never sent to the browser. You can define one in wp-config.php instead:
define( 'SITERYN_WISHLIST_PRO_AI_OPENAI_API_KEY', 'sk-…' ); // Or one key for every built-in provider (not "custom"): define( 'SITERYN_WISHLIST_PRO_AI_API_KEY', 'sk-…' );
Requests & limits
| Setting | Default | Notes |
|---|---|---|
| Request timeout (seconds) | 30 | 5–120 |
| Cache responses for (minutes) | 60 | Identical requests reuse the answer; 0 turns caching off |
| Maximum requests per hour · per day | 100 · 500 | 0 = no limit |
| Log provider errors to the WooCommerce log | Off | Messages and status codes only, never your API key or prompt contents |
AI runs only when you click a button in wp-admin, never on the storefront.
What is sent
Product names, prices, stock and sales figures and wishlist counts (up to 25 products per request), customer-segment head-counts and sometimes your store name. Ask also sends your question and a summary of what the plugin can do; the copy and email drafts send what you type in Anything to mention. Summarise with AI on an estimate request sends the customer's message and the requested product names. Never customer names, emails or orders.
Copy & email drafts
On the AI Studio tab, below the reports. They use the same setup card as the reports.
- Wishlist copy: choose Where it goes (Wishlist page introduction, Empty wishlist message, Add to wishlist button label, Call to action, Promotional message, Wishlist reminder copy or Share wishlist copy), then Draft options.
- Wishlist email: choose a Purpose (Wishlist reminder, Back in stock, Price drop, Promotion or Wishlist to purchase), then Draft email: a subject, preheader, body and call to action built from your most wishlisted products.
- Both have a Tone (Friendly, Professional, Playful, Short and plain, Enthusiastic) and Anything to mention (optional), e.g. a delivery offer.
Drafts are saved nowhere and sent to no one. Check every product, price or offer they mention before using them.
AI insight reports: see AI Studio & activity.