A Firefox extension that detects AI-generated content on LinkedIn posts using pattern matching and linguistic analysis based on Wikipedia's comprehensive guidelines for identifying AI writing.

• Real-time Detection: Automatically analyzes LinkedIn posts as you scroll

• AI Slop Rating: Color-coded badges showing likelihood of AI-generated content

  • 🔴 HIGH AI SLOP (60%+): Very likely AI-generated
  • 🟠 MEDIUM AI SLOP (30-59%): Possibly AI-generated
  • 🟡 LOW AI SLOP (15-29%): Some AI indicators present
  • 🟢 LIKELY HUMAN (<15%): Appears to be human-written

• Detailed Analysis: Click any rating badge to see which specific patterns were detected

• Pattern Categories: Detects 15+ categories of AI writing indicators including:

  • Promotional language and buzzwords
  • Editorializing and superficial analyses
  • Overuse of conjunctives
  • Collaborative communication phrases
  • Knowledge cutoff disclaimers
  • And many more...

• Privacy-Focused: No AI services used - purely rule-based pattern matching

• Customizable: Toggle ratings visibility, auto-analysis, and pattern counts

• Context Menu: Right-click selected text to analyze it for AI patterns

1. Clone or download this repository

   git clone <repository-url>
   cd AI-Slop-Detection

2. Generate Icons (optional - placeholder icons included)

   • Open create_icons.html in your browser
   • Click "Download Icons" to generate proper icon files
   • Place the downloaded icons in the icons/ directory

3. Load Extension in Firefox

   • Open Firefox and go to about:debugging
   • Click "This Firefox" in the left sidebar
   • Click "Load Temporary Add-on"
   • Navigate to the extension folder and select manifest.json

4. Alternative: Load via about:addons

   • Go to about:addons
   • Click the gear icon and select "Debug Add-ons"
   • Click "Load Temporary Add-on" and select manifest.json

The extension would need to be packaged and submitted to Mozilla Add-ons store for regular users.

1. Visit LinkedIn - The extension only works on linkedin.com
2. Browse Posts - AI slop ratings will automatically appear on posts
3. Click Ratings - View detailed analysis of detected patterns
4. Use Popup - Click the extension icon to see stats and settings
5. Right-click Text - Analyze selected text using the context menu

Access settings via the extension popup:

• Show Ratings: Toggle visibility of rating badges
• Auto-Analyze New Posts: Automatically analyze posts as they load
• Show Pattern Count: Display number of detected patterns in badges

The extension uses pattern matching against linguistic markers identified in Wikipedia's "Signs of AI writing" guidelines. It analyzes text for:

• Language & Tone: Promotional language, editorializing, superficial analyses
• Style Issues: Overuse of conjunctives, excessive formatting, curly quotes
• Technical Markers: AI disclosure phrases, knowledge cutoffs, markdown formatting
• Structural Patterns: Negative parallelisms, section summaries, vague attributions

• Each pattern category has a weighted score (1-9 points based on AI confidence)
• Stricter scoring: No caps on pattern multiplication for repeat offenses
• Enhanced text analysis: sentence length, comma usage, conjunctive density
• Thresholds: 60%+ High, 30-59% Medium, 15-29% Low, <15% Human
• Technical markers (AI self-identification) receive highest weights (8-9 points)

• Primary: Firefox (Manifest V2)
• Secondary: Chrome/Edge (with compatibility layer)

• No external connections - all analysis is done locally
• No AI services - purely rule-based detection
• No data collection - nothing is stored or transmitted
• LinkedIn only - extension only activates on LinkedIn

AI-Slop-Detection/
├── manifest.json          # Extension configuration
├── content.js             # Main content script with detection logic
├── popup.html            # Extension popup interface  
├── popup.js              # Popup functionality
├── background.js         # Background script
├── styles.css            # Styling for ratings and modals
├── icons/                # Extension icons
├── create_icons.html     # Icon generator utility
└── README.md            # This file

To add new AI detection patterns:

1. Edit the AI_PATTERNS object in content.js
2. Add new pattern category with regex patterns and weight
3. Higher weights = stronger AI indicators
4. Test thoroughly to avoid false positives

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Test on various LinkedIn posts
5. Submit a pull request

• False Positives: Some human writing may trigger AI patterns
• False Negatives: Sophisticated AI may avoid detected patterns
• LinkedIn Changes: UI updates may break post detection
• Language: Optimized for English content
• Context: Cannot understand semantic meaning, only patterns

This tool provides estimates based on linguistic patterns and should not be considered definitive proof of AI generation. Human judgment is still required for accurate assessment.

MIT License - see LICENSE file for details

Based on the comprehensive AI writing detection guidelines from Wikipedia's "Signs of AI writing" project.