LLMFeeder - Webpage to Markdown for your LLM context!

A browser extension that converts web page content to clean Markdown format and copies it to clipboard with a single click, perfect for feeding content to Large Language Models (LLMs). Available for both Chrome and Firefox.

Website: llmfeeder.fyi

✨ What's New in v2.2.1

• 🖼️ More robust iframe extraction - Same-origin and srcdoc frames are read through the DOM; cross-origin frames keep a link/warning instead of page-to-page HTML messaging
• 🔗 Safer iframe placeholders - Embedded-content links are built with DOM APIs and only for explicit http(s) URLs

Previous Release (v2.2.0)

• 📜 Lazy-Loading Auto-Scroll - Opt-in setting to scroll chat/AI pages (Gemini, ChatGPT, Claude, etc.) before extraction so long conversations aren't truncated (#91)
• 📋 Alternative Copy Mode - Split copy button with a one-click override for Full Page vs Main Content without changing your default setting
• 💻 Better Code Blocks - Generic <pre> elements convert to fenced Markdown code blocks with language detection (#66)
• 🏷️ Store SEO Refresh - Renamed to Web to Markdown for AI; short description targets ChatGPT/Claude/Gemini searches (within Chrome’s 132-char limit)
• 🧪 Unit Tests & CI - Jest test suite and GitHub Actions workflow for PRs
• 🔧 Firefox Icon Fix - Correct 16×16 icon dimensions for Add-ons compatibility

Previous Release (v2.1.0)

• 🗂️ Multi-Tab Support - Convert multiple tabs at once! Select multiple tabs with Ctrl/Cmd+Shift and use Copy All, Download Merged, or Download as ZIP
• 🖱️ Right-Click Context Menus - Quick access to conversion options directly from the right-click menu on any webpage
• 📊 Token Counter - Real-time token count estimation (GPT-4/Claude tokenizer) with configurable context limits (4K to 32K)
• 🔗 Include Links Toggle - New option to include or strip links from Markdown output (reduce token usage when URLs aren't needed)
• 🐛 Debug Mode & Logging - Built-in diagnostic system to help troubleshoot conversion issues
• 🖼️ Cross-Origin Iframe Support - Improved extraction of embedded content from iframes
• 📋 Better Table Conversion - Fixed table-to-markdown conversion for various HTML table formats
• ⭐ Review Prompt - Gentle reminder to rate the extension after 20 successful conversions
• 🔧 Bug Fixes - Fixed race condition in context menu initialization, CSS syntax errors, and removed dead code

Demo

Privacy & Security

LLMFeeder operates as a fully client-side extension with zero backend dependencies. All content processing occurs locally within your browser's execution environment:

• No Remote Data Transmission: The extension performs all operations (content extraction, Markdown conversion, clipboard operations) entirely within your browser's sandbox. No data is ever transmitted to external servers.

• Zero Telemetry: Unlike many extensions, LLMFeeder contains no analytics, tracking, or data collection mechanisms of any kind.

• Minimal Permissions: The extension requests only the permissions strictly necessary for its core functionality (activeTab, clipboardWrite, storage, scripting).

• Verifiable Codebase: Being fully open source, the entire codebase is available for inspection to verify these privacy claims. Users are encouraged to review the source code to confirm the absence of any data exfiltration mechanisms.

This architecture ensures that your content remains exclusively on your device throughout the entire extraction and conversion process.

Features

• Multi-Tab Support: Convert multiple selected tabs at once - merge into single file or download as ZIP
• Lazy-Loading Auto-Scroll: Opt-in scroll pass for chat/AI UIs so long conversations aren't truncated
• Alternative Copy Mode: One-click Full Page / Main Content override from the copy button
• Right-Click Context Menus: Quick conversion access from anywhere on a page
• Token Counter: Real-time token estimation with configurable LLM context limits
• Include Links Toggle: Option to include or strip URLs from links in output (great for reducing token usage)
• Smart Content Extraction: Uses Readability algorithm to focus on main content
• Dark/Light Mode: Toggle between themes with automatic preference persistence
• Enhanced Image Processing: Smart alt text generation with intelligent fallbacks for better image descriptions
• One-Click Simplicity: Single action to process and copy content
• Download as File: Save converted Markdown directly as .md files to your device
• Page Title Integration: Option to include webpage titles in your Markdown output
• Customizable Metadata Format: Template-based system with 6 variables for flexible citation styles
• Debug Mode: Built-in logging system for troubleshooting
• Modern Notifications: Redesigned UI banner with improved accessibility and visual feedback
• LLM-Optimized Output: Clean, structured Markdown perfect for AI consumption
• Customizable: Configure content scope and formatting options
• Keyboard Shortcuts: Quick convert (Alt+Shift+M), download (Alt+Shift+D), and ZIP (Alt+Shift+Z)
• Multi-Browser Support: Works seamlessly on both Chrome and Firefox

Installation

Option 1: Browser Extension Stores (Recommended)

Chrome Web Store

✅ Now Available! Install directly from the Chrome Web Store

Firefox Add-ons

✅ Now Available! Install directly from the Firefox Add-ons Store

Option 2: Direct Download

1. Download the latest release zip file based on your browser.
2. Extract the zip file to a location of your choice

For Chrome:

3. Open Chrome and navigate to chrome://extensions/
4. Enable "Developer mode" by toggling the switch in the top right
5. Click "Load unpacked" and select the extracted extension directory
6. The LLMFeeder extension should now appear in your extensions list
7. Click the puzzle piece icon in Chrome toolbar and pin LLMFeeder for easy access

For Firefox:

3. Open Firefox and navigate to about:debugging#/runtime/this-firefox
4. Click "Load Temporary Add-on..."
5. Select the manifest.json file in the extracted extension directory
6. The LLMFeeder extension should now appear in your add-ons list
7. Note: For permanent installation in Firefox, use the Firefox Add-ons store option when available

Option 3: From Source (Development)

For Chrome

1. Clone this repository:

   git clone git@github.com:jatinkrmalik/LLMFeeder.git
   

2. Build the extension:

• Using the provided build script:

  cd LLMFeeder
  ./scripts/build.sh chrome
  

• Using Makefile (if you have make installed):

  cd LLMFeeder
  make chrome
  

  This will create a Chrome-compatible package in the dist/ directory.

3. Unzip the generated LLMFeeder-Chrome-*.zip file in the dist directory to a directory of your choice
4. Open Chrome and navigate to chrome://extensions/
5. Enable "Developer mode" by toggling the switch in the top right
6. Click "Load unpacked" and select the extension directory from the unzipped package
7. The extension will appear in your Chrome extensions list

For Firefox

1. Clone this repository:

   git clone git@github.com:jatinkrmalik/LLMFeeder.git
   

   (If you've already cloned, ensure you are on the desired branch/commit.)

2. Navigate to the cloned directory:

   cd LLMFeeder
   

3. Build the extension:

• Using the provided build script:

  ./scripts/build.sh firefox
  

• Using Makefile (if you have make installed):

  make firefox
  

  This will create a Firefox-compatible package in the dist/ directory.

4. Unzip the generated LLMFeeder-Firefox-*.zip file in the dist directory to a directory of your choice
5. Open Firefox and navigate to about:debugging#/runtime/this-firefox
6. Click "Load Temporary Add-on..."
7. Select the manifest.json file inside the extension directory from the unzipped package
8. The extension will appear in your Firefox extensions list

Usage

1. Navigate to any web page you want to convert to Markdown
2. Click the LLMFeeder icon in your browser toolbar
3. Configure your preferences (content scope, include page title, etc.)
4. Choose your action:
   • "Convert & Copy" - Process content and copy to clipboard
   • "Download as .md" - Save content directly as a Markdown file
5. Paste the clipboard content into your LLM tool or use the downloaded file

Keyboard Shortcuts

• Open Extension Popup: Alt+Shift+L (Windows/Linux) or ⌥⇧L (Mac)
• Convert & Copy without Opening Popup: Alt+Shift+M (Windows/Linux) or ⌥⇧M (Mac)
• Download as Markdown File: Alt+Shift+D (Windows/Linux) or ⌥⇧D (Mac)
• Download Multiple Tabs as ZIP: Alt+Shift+Z (Windows/Linux) or ⌥⇧Z (Mac)

Customizing Shortcuts

Users can customize keyboard shortcuts by following these steps:

• Chrome: Go to chrome://extensions/shortcuts in your browser
• Firefox: Go to about:addons → Extensions → ⚙️ (Gear icon) → Manage Extension Shortcuts

Settings

• Content Scope:

  • Main article content only (default)
  • Full page content
  • Selected text only

• Content Options:

  • Include page title in output
  • Enhanced image alt text generation

• Formatting Options:

  • Preserve table formatting
  • Include/exclude images
  • Include/exclude links (strip URLs to reduce token usage)

• Metadata Format:

  • Toggle metadata inclusion on/off
  • Customize metadata template with 6 variables: {title}, {url}, {date}, {author}, {siteName}, {excerpt}
  • Reset to default format with one click
  • Predictable WYSIWYG formatting

• Output Options:

  • Copy to clipboard
  • Download as .md file

Project Structure

Below is a visual representation of the project structure to help developers understand the codebase:

LLMFeeder/
│
├── dist/                     # Distribution packages directory
│   ├── LLMFeeder-Chrome-*.zip # Chrome package
│   ├── LLMFeeder-Firefox-*.zip # Firefox package
│   └── LLMFeeder-Source-*.zip # Source package
│
├── extension/                 # Browser extension directory
│   ├── icons/                 # Extension icons
│   │   ├── icon16.png         # 16x16 icon
│   │   ├── icon48.png         # 48x48 icon
│   │   └── icon128.png        # 128x128 icon
│   │
│   ├── libs/                  # Third-party libraries
│   │   ├── readability.js     # Mozilla's Readability for content extraction
│   │   ├── turndown.js        # HTML to Markdown conversion
│   │   └── browser-polyfill.js # Browser compatibility layer
│   │
│   ├── manifest.json          # Extension configuration and metadata
│   ├── popup.html             # UI for the extension popup
│   ├── popup.js               # Controls popup behavior and user interactions
│   ├── styles.css             # Styling for the popup UI
│   ├── content.js             # Content script that runs on web pages
│   └── background.js          # Background script for keyboard shortcuts
│
├── scripts/                   # Build and utility scripts
│   └── build.sh               # Main build script
│
├── testbench.html             # Test bench page for extension testing
├── iframe-content-test.html   # Nested iframe content for testing
│
├── .gitignore                 # Git ignore rules
└── README.md                  # Project documentation

Component Descriptions

• manifest.json: Defines extension metadata, permissions, and configuration
• popup.html/js/css: Creates the user interface when you click the extension icon
• content.js: Contains the core functionality to extract and convert web content to Markdown
• background.js: Handles keyboard shortcuts and global extension functionality
• readability.js: Mozilla's library that identifies and extracts the main content from a webpage
• turndown.js: Converts HTML to Markdown with configurable options
• browser-polyfill.js: Provides compatibility layer for Chrome and Firefox extension APIs

Data Flow

1. User triggers conversion (via popup UI or keyboard shortcut)
2. Request is sent from popup.js or background.js to content.js
3. content.js extracts content using readability.js
4. Enhanced image alt text generation processes images with smart fallbacks
5. Optional page title is prepended if enabled
6. Content is converted to Markdown using turndown.js
7. Output is either copied to clipboard or downloaded as .md file
8. Cross-browser notification provides success feedback

Browser Compatibility

LLMFeeder is designed to work on all modern browsers with complete feature parity:

• Chrome/Chromium-based browsers: Fully supported (v80+)
• Firefox: Fully supported (v109+)

Tech Stack

• Content Extraction: Mozilla's Readability.js
• Markdown Conversion: Turndown.js
• Extension Framework: Web Extensions API (Chrome Manifest V3, Firefox Manifest V2 compatibility)
• Browser Compatibility: Custom polyfill for cross-browser support

Build Instructions

Prerequisites

• zip installed (required for packaging)
• jq installed (optional but recommended for JSON processing)
• For Make approach: GNU Make (usually pre-installed on Linux/macOS)

---

Using Scripts

To build the extension packages for distribution:

1. Run the build script:

   ./scripts/build.sh
   

This will create three packages in the dist/ directory:

• LLMFeeder-Chrome-v1.0.1.zip - Chrome-compatible package
• LLMFeeder-Firefox-v1.0.1.zip - Firefox-compatible package
• LLMFeeder-Source-v1.0.1.zip - Source code package

You can also build specific packages:

# Build only Chrome package
./scripts/build.sh chrome

# Build only Firefox package
./scripts/build.sh firefox

# Build only source package
./scripts/build.sh source

# Specify version number
./scripts/build.sh --version 1.1.0 all

Using Make

To build the extension packages for distribution:

• Common Targets

make all           # Build all (chrome, firefox, source)
make chrome        # Build Chrome only
make firefox       # Build Firefox only
make source        # Build source zip only
make full-build    # Full build via ./scripts/build.sh

• Versioned Builds

make versioned-chrome version=1.0.0
make versioned-full version=1.0.0
make versioned-firefox version=1.0.0
make versioned-source version=1.0.0

Note: ❗ These require a version=x.y.z flag or will throw an error.

• Utilities

make clean         # Remove dist/ artifacts
make help          # Show help message

Testing

Unit Tests

The project uses Jest for unit testing. Tests live in the tests/ directory, with Chrome extension APIs mocked in tests/setup.js.

1. Install dependencies (requires Node.js 20+):

   npm install
   

2. Run the test suite:

   npm test
   

Jest prints a pass/fail summary per test file; a non-zero exit code means at least one test failed. The suite also runs automatically in CI (GitHub Actions) on every push to main and on every pull request — see .github/workflows/test.yml.

To add tests for a new module, export it for Node at the bottom of the source file (see extension/token-counter.js for the pattern):

if (typeof module !== 'undefined' && module.exports) {
  module.exports = MyModule;
}

Then create tests/my-module.test.js following the structure of tests/token-counter.test.js.

A test bench page (testbench.html) is included to verify extension functionality manually:

Running the Test Bench

1. Start a local HTTP server:

   python3 -m http.server 8080
   

2. Open http://localhost:8080/testbench.html in your browser

3. Test the extension with different content types:

   • Main Content Extraction: Article content with Readability
   • Iframe Extraction: Same-origin, srcdoc, and cross-origin iframes
   • Table Conversion: Standard HTML tables and GitHub-style tables
   • Empty Cells: Tables with missing data

Test Coverage

Test Case	Description	Expected Result
Test 1	Main article content	Extracted with proper formatting
Test 2	srcdoc iframe	Content extracted (same-origin)
Test 3	Nested iframe	Content extracted (same-origin)
Test 4	Cross-origin iframe	Warning message with link placeholder
Test 5	Sandboxed iframe	May or may not extract (browser-dependent)
Test 6	Standard HTML table	Markdown table with `
Test 7	GitHub-style table	Markdown table without <thead>
Test 8	Empty cells	Empty cells handled correctly

Testing Tips

• Enable Debug Mode in options for detailed logs
• Use Full Page scope to test iframe extraction
• Enable Preserve Tables to test table conversion
• Test with both "Main Content" and "Full Page" scopes

License

MIT

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

Our Wonderful Contributors 💪

Thanks a lot for spending your time helping LLMFeeder grow.🍻

Star Chart