diff --git a/CLAUDE.md b/CLAUDE.md
new file mode 100644
index 0000000..3543ed1
--- /dev/null
+++ b/CLAUDE.md
@@ -0,0 +1,164 @@
+# CLAUDE.md
+
+This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
+
+## Project Overview
+
+This is a Chromium browser extension for WPlace (wplace.live), a collaborative pixel art canvas website. The extension provides automation tools to farm pixels, place images, repair artwork, extract artwork, and manage multiple accounts.
+
+Key features include:
+- Auto Farm: Automatically farms pixels with configurable charge thresholds
+- Auto Image: Places images on the canvas with overlay support
+- Auto Repair: Automatically repairs existing artwork
+- Art Extractor: Extracts artwork areas and saves them as JSON
+- Multi-Account Manager: Switch between and manage multiple WPlace accounts
+- Script Manager: Neon-themed interface to launch automation scripts
+
+## Repository Structure
+
+```
+Extension/
+├── background.js              # Background service worker (cookie/account management, script execution)
+├── content-script.js          # Content script (injects UI button into wplace.live)
+├── Script-manager.js          # Main script launcher (neon-themed UI)
+├── manifest.json              # Extension manifest
+├── icons/                     # Extension icons
+├── lang/                      # Language translation files (JSON)
+├── themes/                    # CSS theme files
+├── popup/                     # Extension popup UI
+│   ├── popup.html             # Popup interface
+│   ├── popup.js               # Popup logic (account management)
+│   └── popup.css              # Popup styling
+└── scripts/                   # Automation scripts
+    ├── Auto-Farm.js           # Pixel farming automation
+    ├── Auto-Image.js          # Image placement automation
+    ├── Auto-Repair.js         # Artwork repair automation
+    ├── Art-Extractor.js       # Artwork extraction tool
+    ├── Acc-Switch.js          # Account switching utilities
+    ├── image-processor.js     # Image processing utilities
+    ├── overlay-manager.js     # Canvas overlay management
+    ├── token-manager.js       # Turnstile token management
+    └── utils-manager.js       # Shared utility functions
+```
+
+## Architecture Overview
+
+### Extension Components
+
+1. **Manifest V3 Extension**: Uses modern Chrome extension architecture
+2. **Background Service Worker**: Handles account management, cookie handling, and script execution
+3. **Content Script**: Injects a button into wplace.live for launching automation scripts
+4. **Popup UI**: Provides account management and script selection interface
+5. **Script Manager**: Neon-themed launcher for all automation scripts
+6. **Individual Scripts**: Separate modules for different automation tasks
+
+### Key Architectural Patterns
+
+1. **Message Passing**: Content script ↔ Background script ↔ Page context
+2. **Resource Loading**: Background script pre-loads themes and language files, injects them into scripts
+3. **MAIN World Execution**: Scripts run in MAIN world context to interact with wplace.live APIs
+4. **Dependency Injection**: Complex scripts inject utility managers (image-processor, overlay-manager, etc.)
+5. **Global Managers**: Singleton utility managers available as global objects (window.globalUtilsManager, etc.)
+
+### Data Flow
+
+1. User clicks extension button on wplace.live
+2. Content script sends message to Background script
+3. Background script:
+   - Loads script dependencies (managers, themes, languages)
+   - Fetches script content
+   - Injects resources and executes in MAIN world
+4. Script runs in page context with full access to wplace.live APIs
+5. UI elements (panels, overlays) are injected into the page DOM
+
+### Account Management
+
+Accounts are stored in chrome.storage.local with:
+- `accounts`: Array of authentication tokens
+- `infoAccounts`: Detailed account information (name, charges, droplets, etc.)
+- Background script handles cookie setting/getting for account switching
+
+## Development Guidelines
+
+### No Build Process Required
+
+This is a browser extension with no build process. Development happens directly with the source files:
+
+1. Make code changes
+2. Reload extension in Chrome (chrome://extensions/)
+3. Test on wplace.live
+
+### File Modification Guidelines
+
+1. **Never modify** `manifest.json` unless adding new permissions or resources
+2. **Prefer editing** existing language JSON files over adding new ones
+3. **Script files** can be modified directly - they're injected as-is
+4. **CSS files** are loaded and injected by the background script
+
+### Extension Loading for Development
+
+1. Open Chrome/Edge and navigate to `chrome://extensions/`
+2. Enable "Developer mode"
+3. Click "Load unpacked"
+4. Select the `Extension` directory
+5. Test on https://wplace.live
+
+### Testing Changes
+
+1. Make changes to source files
+2. Reload the extension in Chrome extensions page
+3. Refresh wplace.live
+4. Click the extension button to test functionality
+
+### Common Development Tasks
+
+#### Adding a New Script
+1. Create script file in `Extension/scripts/`
+2. Add entry to `AVAILABLE_SCRIPTS` in:
+   - `Extension/Script-manager.js`
+   - `Extension/popup/popup.js`
+3. Update manifest if new permissions needed
+
+#### Adding a New Language
+1. Create JSON file in `Extension/lang/`
+2. Add to language loading in `Extension/background.js`
+3. No code changes needed - automatically loaded
+
+#### Adding New UI Elements
+1. CSS is automatically loaded from `Extension/themes/`
+2. For popup changes, modify files in `Extension/popup/`
+3. For in-page UI, modify individual script files
+
+## Key APIs and Integration Points
+
+### WPlace Integration
+- Canvas interaction through simulated mouse events
+- API calls to `https://backend.wplace.live/` for account data
+- DOM manipulation to add UI elements
+
+### Chrome Extension APIs
+- `chrome.storage.local` for account data persistence
+- `chrome.cookies` for authentication token management
+- `chrome.scripting` for injecting scripts into MAIN world
+- `chrome.runtime` for message passing
+
+### Global Utility Managers
+- `window.globalUtilsManager` - Shared utility functions
+- `window.globalImageProcessor` - Image processing functions
+- `window.globalOverlayManager` - Canvas overlay functions
+- `window.globalTokenManager` - Turnstile token functions
+
+## Debugging
+
+- All scripts log to browser console with colored output
+- Extension background console available at `chrome://extensions/`
+- Console logs include detailed execution information
+- Use browser dev tools to inspect injected UI elements
+
+## Multi-Language Support
+
+The extension supports multiple languages through JSON files in `Extension/lang/`. Text is automatically translated based on user's browser language. Translation functions are available through `window.getLanguage()` and `t()` functions in utils manager.
+
+## Themes
+
+Visual styling is handled through CSS files in `Extension/themes/`. The neon theme is the primary theme used throughout the extension. Themes are loaded by the background script and injected into pages.
\ No newline at end of file
diff --git a/Extension/scripts/Auto-Repair.js b/Extension/scripts/Auto-Repair.js
index 4f6abed..ad0ba35 100644
--- a/Extension/scripts/Auto-Repair.js
+++ b/Extension/scripts/Auto-Repair.js
@@ -106,6 +106,7 @@
     region: null,
     paintWhitePixels: true,
     paintTransparentPixels: true, // Changed to true to fix transparent pixel detection
+    scanBottomUp: false, // New state for bottom-up scanning
     autoRepairEnabled: false,
     autoRepairInterval: 30,
     autoRepairTimer: null,
@@ -1671,14 +1672,19 @@
       Utils.addDebugLog('Using restored data - skipping tile wait for damage scan', 'info');
     }
 
-    Utils.addDebugLog(`Scanning ${width}x${height} image for damage (transparent detection: ${state.paintTransparentPixels})...`, 'info');
+    Utils.addDebugLog(`Scanning ${width}x${height} image for damage (transparent detection: ${state.paintTransparentPixels}, bottom-up: ${state.scanBottomUp})...`, 'info');
 
     let scannedPixels = 0;
     let transparentPixelsDetected = 0;
     let wrongColorPixelsDetected = 0;
     let lastProgressUpdate = Date.now();
-    
-    for (let y = 0; y < height; y++) {
+
+    // Determine scan direction
+    const yStart = state.scanBottomUp ? height - 1 : 0;
+    const yEnd = state.scanBottomUp ? -1 : height;
+    const yStep = state.scanBottomUp ? -1 : 1;
+
+    for (let y = yStart; y !== yEnd; y += yStep) {
       for (let x = 0; x < width; x++) {
         if (state.stopFlag) break;
 
@@ -1807,10 +1813,12 @@
       }
 
       if (state.autonomousMode && Date.now() - lastProgressUpdate > 5000) {
-        Utils.addDebugLog(`Scan progress: ${y}/${height} rows (${Math.round((y / height) * 100)}%), found ${damagedPixels.length} damaged (${transparentPixelsDetected} transparent, ${wrongColorPixelsDetected} wrong color)`, 'info');
+        const progressY = state.scanBottomUp ? (height - 1 - y) : y;
+        Utils.addDebugLog(`Scan progress: ${progressY}/${height} rows (${Math.round((progressY / height) * 100)}%), found ${damagedPixels.length} damaged (${transparentPixelsDetected} transparent, ${wrongColorPixelsDetected} wrong color)`, 'info');
         lastProgressUpdate = Date.now();
       } else if (!state.autonomousMode && y % 10 === 0) {
-        Utils.addDebugLog(`Scan progress: ${y}/${height} rows (${Math.round((y / height) * 100)}%)`, 'info');
+        const progressY = state.scanBottomUp ? (height - 1 - y) : y;
+        Utils.addDebugLog(`Scan progress: ${progressY}/${height} rows (${Math.round((progressY / height) * 100)}%)`, 'info');
       }
     }
 
@@ -2830,6 +2838,14 @@
                 <span class="batch-control-label">🔧 Repair Transparent Pixels</span>
               </label>
             </div>
+
+            <!-- Bottom-Up Scanning Control -->
+            <div class="batch-control-group">
+              <label style="display: flex; align-items: center; cursor: pointer;">
+                <input type="checkbox" id="scanBottomUp" ${state.scanBottomUp ? 'checked' : ''} class="neon-checkbox">
+                <span class="batch-control-label">⬇️ Scan Bottom-Up (Y-axis)</span>
+              </label>
+            </div>
           </div>
           
           <div style="margin-top: 10px; font-size: 8px;" class="neon-text info">
@@ -3192,6 +3208,12 @@
       Utils.addDebugLog(`Transparent pixel repair ${state.paintTransparentPixels ? 'enabled' : 'disabled'}`, 'info');
     });
 
+    // Bottom-up scanning toggle
+    document.getElementById('scanBottomUp').addEventListener('change', (e) => {
+      state.scanBottomUp = e.target.checked;
+      Utils.addDebugLog(`Bottom-up scanning ${state.scanBottomUp ? 'enabled' : 'disabled'}`, 'info');
+    });
+
     // Clear debug
     document.getElementById('clearDebugBtn').addEventListener('click', () => {
       state.debugLogs = [];
@@ -3223,13 +3245,20 @@
       data.version = '3.0';
       Utils.addDebugLog('Migrated to autonomous features v3.0', 'info');
     }
+
+    // Migration for bottom-up scanning feature
+    if (data.version === '3.0') {
+      data.state.scanBottomUp = data.state.scanBottomUp ?? false;
+      data.version = '3.1';
+      Utils.addDebugLog('Migrated to bottom-up scanning feature v3.1', 'info');
+    }
     
     return data;
   }
 
   // Main initialization function
   async function initialize() {
-    Utils.addDebugLog('Starting WPlace Autonomous Repair Tool v3.0...', 'info');
+    Utils.addDebugLog('Starting WPlace Autonomous Repair Tool v3.1...', 'info');
     
     // Create UI first
     createUI();