AI Summary Generation 1.0.5

AI Summary is a content summary generation add-on designed for XenForo forums, capable of automatically analyzing thread content and generating concise AI summaries with tags extraction.

Core Features

AI Summary Generation: Automatically analyzes thread post content to generate summaries
Tags Extraction: Identifies and displays important tags/keywords from content
Post Links: Optional clickable post links included in summaries
Like/Dislike System: Users can rate AI-generated summaries
Permission Control: Flexible user group permission management with rate limiting
Theme Selection: Choose from preset themes (Default, Light, Dark)
Performance Tracking: Track generation time, post count, etc. to gain thread insights and assess summary quality.

Front-end example can be seen here:

Administrator Settings

Setting Location:

• Admin Panel → Setup → Options → AI Summary Settings
• Admin Panel → Add-ons → AI Summary → click the settings icon → Options

License Key:

• Description: Enter your valid license key for the add-on
• Required: Yes - (available updates will be downloadable with a valid license)
• Format: Standard license key format provided upon purchase

Enable AI Summary Widget

• Options: On/Off
• Description: When enabled, the AI Summary widget will be displayed on thread pages. When disabled, the widget will be hidden from all users.

Display Nodes (Forums)

• Description: Select the forums where the AI Summary widget should be shown. Leave empty to show in all forums.

Visible User Groups

• Description: Select all user groups that can see the AI Summary widget. If no groups are selected, all users can see the widget by default.

Theme Colors

• Theme Options: Choose from Default(Auto-detection), Light, Dark
• Description: Choose the color theme for the AI Summary widget. Default automatically adapts to the current forum theme, Light uses a light color scheme, and Dark uses a dark color scheme.
• Note: Auto-detection only works for default XenForo themes.

Enable Permission Control

• Options: On/Off
• Description: When enabled, user group rate limits are enforced

Group Rate Limits

• Configuration: Per-user-group rate limiting

Access Levels:

• Not Allowed: Completely disabled for the group
• Unlimited: No restrictions
• Limited: Configurable rate limits

Rate Limits:

• Per Hour: Maximum requests per hour
• Per Day: Maximum requests per day
• Per Month: Maximum requests per month

Permission Scope:

• Generation Only: Only initial summary generation
• Regeneration Only: Only summary regeneration
• Both: Both generation and regeneration

Deployment Type

Options:

• API Provider: Use remote AI services (OpenRouter, OpenAI, Google AI, Anthropic, etc.)
• Local Deployment (Ollama): Use local AI services

Default: API Provider

• Note: Local deployment requires technical setup and server resources

API URL

• Description: Set the API URL address for AI services

API URL Examples:

• OpenRouter: https://openrouter.ai/api/v1/chat/completions
• OpenAI: https://api.openai.com/v1/chat/completions
• Anthropic: https://api.anthropic.com/v1/messages
• Google AI: https://generativelanguage.googleapis.com/v1beta/models/gemini-2.5-flash:generateContent
• Local Ollama Deployment: http://your-server-ip:11434

API Key

• Description: Enter your API key for the selected service
• Local Deployment: Optional (not required for Ollama)
• API Provider: Required for most services
• Default Key: Leave it empty to use default free OpenRouter API key

Note: Default API Key can easily reach rate limits. Highly recommend to use your own API Keys.

• Get OpenRouter API Keys at https://openrouter.ai/docs/api-reference/authentication
• Or visit other API Keys providers (OpenAI, Google AI, Anthropic, etc.)

AI Model

Free Models (OpenRouter):

• deepseek/deepseek-r1-0528:free (Default recommended)
• google/gemini-2.0-flash-exp:free
• qwen/qwen3-14b:free
• mistralai/mistral-small-3.2-24b-instruct:free
• shisa-ai/shisa-v2-llama3.3-70b:free
• microsoft/mai-ds-r1:free

Premium Models: Available with custom API keys

• Local Models: Any Ollama-compatible model name
• Custom Models: Support for other model names (compatibility not guaranteed)

AI Prompt

Prompt Mode: Choose between default and custom prompt modes

• Custom Prompt: Override default AI prompts for summary generation
• Custom Prompt Examples:

Without post links:

Task:
Please analyze the following forum discussion content and provide a response STRICTLY in English:

CONTENT FORMAT: The content includes post markers like [123]...[/123] to identify individual posts. Ignore these markers in your analysis and focus only on the actual content.

ANALYSIS TASK:
1. Three key words separated by commas - choose the most important topics
IMPORTANT: Each keyword must start with a capital letter (e.g., 'Product Review, Price Analysis, Quality Assessment')
2. A concise but comprehensive summary - be brief yet informative, capture the main points and conclusions

IMPORTANT: Your entire response must be in English, regardless of the input language. Keep your summary concise but don't sacrifice important details. Do NOT include or reference the post markers [numbers] in your output.

Format your response as:
Keywords: [Keyword1, Keyword2, Keyword3]
Summary: [concise but comprehensive summary in English]
Content to analyze:
{CONTENT}

With post links:

Task:
Read the forum content below. Post IDs appear as [123], [456], etc. Write your response in concise English.

Config:
BASE_URL = https://[your-domain.com/your-subpath]

Rules for post references (CRITICAL):

Render every referenced post as EXACTLY this HTML:
^{<a href="BASE_URL/posts/POST_ID" target="blank" class="ai-summary-link">POST_ID</a>}
- Replace POST_ID with the numeric ID from markers (e.g., [123] → 123).
Do NOT include any “#” before POST_ID in the link text.
Use absolute URL only (prefix with BASE_URL).
Allowed HTML tags: <sup>, <a>, <br>. Allowed attributes: href, target, class.
Do NOT use onclick or any other attributes.
Output format:
Keywords: [Keyword1, Keyword2, Keyword3]
Summary: concise English summary that naturally includes the superscript anchors above.
Content to analyze:
{CONTENT}


Enable Post Links

• Options: On/Off
• Description: When enabled, clickable post links are included in summaries
• Note: Enabling will increase API usage and processing time

Summary Scope

Options:

• All posts in thread: Analyzes the entire thread
• Posts from Specific Page: Only analyzes specific pages

Description: Choose the content range for summary generation

Start Page Number
Range: 0-10000
Description: Specify the starting page for summary generation, 0 represents the last page

End Page Number
Range: 0-10000
Description: Specify the ending page for summary generation, 0 represents the last page

Max Posts Per Thread
Default Value: 100
Range: 1-10000
Description: Limits the maximum number of posts per thread used for summary generation

Maximum Content Length
Default Value: 20000
Range: 1000-200000
Description: Limits the maximum number of characters sent to AI for processing
Note: Setting too high may exceed AI model input limits

User Guide

Generating AI Summaries

Step 1: Access Thread

• Navigate to any thread in your forum
• Find the " Generate AI Summary" button at the top of the thread

Step 2: Generate Summary

• Click the " Generate AI Summary" button
• Wait for generation to complete (displays "Generating... Please wait." with progress)
• The summary will appear at the top of the thread after generation

Step 3: View Summary

• Summary Content: Displays AI-generated thread summary
• Tags: Shows extracted important keywords/tags
• Post Links: If enabled, includes clickable post links
• Generation Info: Shows generation time etc.

Using Like/Dislike Function

Voting Operations

• Like: Click the button to like the summary
• Dislike: Click the button to dislike the summary
• Cancel Vote: Click the voted button again to cancel
• Change Vote: Click a different button to change your vote

Vote Status

• Highlight Display: Currently voted buttons are highlighted
• Count Update: Vote counts update in real-time
• User Isolation: Each user's vote status is independent

Regenerating Summaries:

Regeneration

• Click the " Regenerate AI Summary" button
• Wait for new summary generation to complete
• All vote records are automatically reset

Notes

• Regeneration overwrites the original summary
• All vote records are cleared
• New summary starts collecting votes again

Troubleshooting

Common Issues
Issue 1: AI Summary Cannot Generate

Possible Causes:

• Invalid or missing license key
• Deployment type configuration error
• API Key not set or incorrect
• API URL configuration error
• Network connection issues
• AI model selection error
• Rate limits exceeded

Solutions:

• Verify license key is valid and properly configured
• Confirm deployment type is selected correctly
• Check if API Key is set correctly
• Verify if API URL is correct
• Test network connection
• Confirm if AI model is available
• Check user group rate limits
• Use recommended default settings for testing

Issue 2: Permission Control Abnormal

Possible Causes:

• User group settings error
• Rate limit configuration issues
• Cache issues
• Settings not saved properly

Solutions:

• Check user group rate limit settings
• Verify rate limit configuration
• Clear XenForo cache
• Re-save settings

Issue 3: API Service Errors

Possible Causes:

• Invalid API key
• API service down
• Rate limits exceeded
• Model not available
• Network connectivity issues

Solutions:

• Verify API key is valid and active
• Check API service status
• Monitor API usage and limits
• Try different AI models
• Check network connectivity

Debug Tools

Browser Developer Tools

• Open F12 developer tools
• View error information in Console tab
• Check API requests in Network tab
• Verify DOM structure in Elements tab

Server Logs

• Check XenForo error logs
• View PHP error logs
• Monitor API call logs
• Verify database query logs

Add-on Diagnostics

• Check license validation status
• Verify API connectivity
• Test AI model availability
• Review permission settings

Contact Support

• Website: https://quantivedynamics.com/
• Email: support@quantivedynamics.com

When reporting issues, please provide:

• Detailed problem description
• Reproduction steps
• Error screenshots or logs
• Browser and system information
• Server environment information
• Add-on configuration details

Updates and Maintenance
Regular Updates

• Feature Updates: New functionality and improvements
• Bug Fixes: Stability and compatibility improvements
• Security Updates: Security patches and vulnerability fixes

Maintenance

• License Renewal: Annual license renewal required
• API Key Management: Monitor and update API keys as needed
• Performance Monitoring: Monitor API usage and costs
• Backup: Regular database backups recommended