How Can We Help?
Plugin Settings: Complete Reference Guide
This is the complete reference for every setting in AIWU. If you can’t find a setting, use your browser’s Ctrl+F to search this page.
Jump to section: General · AI Providers · Chatbot · Training · Workflows · MCP · Advanced
Before You Start
Helpful background:
- Which AI provider is right for you (Choose an AI Provider)
- What each AIWU feature does (Feature Guide)
How to Access Plugin Settings
All settings are in WordPress Admin → AI Copilot. Each tab in the left menu corresponds to a section below.
General Settings
Location: AI Copilot → Settings → General
| Setting | What it does | Default |
|---|---|---|
| Plugin Language | Admin interface language | Auto-detect from WordPress |
| AI Response Language | Language for AI-generated text. “Auto” detects the user’s query language. | Auto |
| Logging Level | What gets saved to logs. Debug = everything. Error = problems only. | Error |
| Log Retention | How many days to keep logs before auto-delete | 30 days |
| Usage Statistics | Track token usage per provider, per workflow, per user | Enabled |
| Monthly Usage Alerts | Email notification when token usage exceeds a threshold | Disabled |
AI Providers
Location: AI Copilot → Settings → AI Providers
This is where you enter API keys for each AI service. AIWU supports multiple providers simultaneously — you can use different providers for different tasks.
Provider Settings (same structure for each provider)
| Setting | What it does |
|---|---|
| API Key | Your API key from the provider’s dashboard. Never share this. |
| Default Model | Which model to use when this provider is selected |
| Max Tokens (response) | Maximum length of AI response. Higher = longer responses + more API cost. |
| Temperature | Creativity level. 0 = precise/deterministic. 1 = creative/varied. For factual content, use 0.2–0.4. |
| Request Timeout | Seconds to wait before giving up. Increase for slow connections. |
| Rate Limit (per minute) | Max API calls per minute. Set below your plan’s limit to avoid errors. |
💡 Provider selection order: When you set a “Default AI Provider” in General Settings, that provider is used for chatbot conversations. Workflows can override this per-action. Training (embeddings) uses a separate provider selection.
Chatbot Settings
Location: AI Copilot → Chatbot
Appearance Tab
| Setting | What it does |
|---|---|
| Chat Widget Position | Where the chat bubble appears (bottom-right, bottom-left, etc.) |
| Primary Color | Main color for buttons and the header bar |
| Font Size | Text size inside the chat window |
| Chat Window Height | Pixel height of the open chat window |
| Mobile Behavior | Show as full-screen on mobile, or keep as widget |
| Custom CSS | Advanced: inject custom styles directly into the chat widget |
Behavior Tab
| Setting | What it does |
|---|---|
| Welcome Message | First message the bot sends when a user opens chat |
| Suggested Questions | Clickable starter questions shown below the welcome message |
| Memory / Session Length | How many messages to include in conversation history for context |
| Typing Indicator | Show “…” animation while AI generates response |
| Response Streaming | Show response word-by-word as it generates, instead of waiting for the full response |
| Show Sources | Show which knowledge base articles the AI used to generate its answer |
| Collect Email | Ask for email before starting chat (for lead generation) |
| Offline Mode Message | Message shown when the AI service is unreachable |
System Prompt Tab
| Setting | What it does |
|---|---|
| Bot Name | Name displayed in the chat header |
| System Prompt | Instructions that define the bot’s persona, tone, and knowledge scope. This is sent with every message. |
| Topics to Avoid | Comma-separated list of topics the bot should decline to discuss |
| Escalation Phrase | If a user says this phrase, route them to human support |
WooCommerce Tab (appears when WooCommerce is active)
| Setting | What it does |
|---|---|
| Product Search in Chat | Allow AI to search your product catalog during conversation |
| Add to Cart from Chat | Allow the AI to add products to the user’s cart directly from chat |
| Order Status Lookup | Allow users to check their order status by providing order number |
| Show Product Cards | Display product image + price as a card, not just text, in chat responses |
| Max Products in Response | Maximum number of products to suggest in a single chat message |
Training Settings
Location: AI Copilot → Training → Settings
| Setting | What it does |
|---|---|
| Embeddings Provider | Which AI service generates your embeddings. OpenAI text-embedding-3-small is recommended for most sites. |
| Embeddings Model | Specific model used for generating embeddings |
| Vector Database | Where embeddings are stored: MySQL, Pinecone, or Weaviate. See Vector DB Comparison. |
| Auto-Train on Publish | Automatically generate embeddings when you publish or update a post |
| Batch Size | How many items to process per batch during bulk training. Lower = slower but less server load. |
| Search Relevance Threshold | Minimum similarity score (0–1) for a result to be returned. Higher = stricter matching. |
| Hybrid Search Weight | Balance between keyword search and semantic search. 0.5 = equal weight. |
Workflow Settings
Location: AI Copilot → Workflows → Settings
| Setting | What it does |
|---|---|
| Max Execution Time | Seconds a workflow can run before being terminated. Increase for long workflows. |
| Default Retry Count | How many times to retry a failed action before marking workflow as failed |
| Error Notification Email | Email address for workflow failure alerts |
| Async Execution | Run workflows in background (non-blocking). Recommended for workflows with multiple AI calls. |
| Workflow Log Retention | Days to keep workflow execution logs |
| Concurrent Workflow Limit | Maximum workflows running at the same time. Increase if workflows queue up. |
MCP Settings
Location: AI Copilot → MCP → Settings
| Setting | What it does |
|---|---|
| Enable MCP Server | Master on/off switch for the MCP module |
| Authentication Method | Token (URL parameter) or OAuth 2.1 (recommended for Claude.ai) |
| MCP Token | Your access token for token-based auth. Click Regenerate to create a new one. |
| Enable Remote MCP (OAuth) | Allow OAuth 2.1 flow for browser-based AI clients like Claude.ai |
| Allowed Tools | Which MCP tools are available: Posts, WooCommerce, Comments, Workflows, Media, Users |
| Rate Limit | Maximum MCP requests per minute per connected client |
Advanced Settings
Location: AI Copilot → Settings → Advanced
⚠️ Warning: Advanced settings can affect site performance and stability. Only change these if you understand the consequences.
| Setting | What it does |
|---|---|
| Debug Logging | Log detailed debug information (verbose). Only enable temporarily when troubleshooting. |
| API Request Caching | Cache identical API requests for a set period to reduce API costs. Default: 1 hour. |
| PHP Memory Override | Attempt to increase PHP memory limit for AI processing. Requires server-level support. |
| Custom Endpoint Prefix | Change the default /aiwu/ REST API prefix if it conflicts with other plugins |
| Disable Emoji in Responses | Strip emoji characters from AI responses |
| Data Retention: Conversations | Days to keep chat conversation logs |
| Data Retention: Embeddings | What happens to embeddings when a post is deleted: Keep, Archive, or Delete |
| Reset Plugin Data | Delete all AIWU data (conversations, embeddings, settings). Irreversible. |
What’s Next
- 🔐 Security settings explained: Security & Privacy: How AIWU Handles Your Data
- 🤖 Set up your first chatbot: Set Up First AI Chatbot
- 🔌 API keys explained: REST API Quick Start
Last verified: AIWU v.4.9.2 · Updated: 2026-02-25