Seer is a Chrome extension that highlights webpages with unknown words and helps you identify good targets for mining in Anki.
️This extension is still under heavy development and may not work as expected. No functionality or claims of stability are made. This is mostly for personal use at the moment, but I figured I'd share it with the community.
This extension is scoped as a vocabulary tracker and a page highlighter. It is just meant to let you easily evaluate the difficulty of content and identify words to mine more easily. Dictionary lookups should be done with a tool like Yomitan. If you're mining video content, feel free to use ASBPlayer.
-
Prerequisites
- Node.js (version 16 or higher)
- Chrome browser
-
Clone and Build
git clone https://github.com/simonbukin/seer.git cd seer npm install npm run build -
Load Extension in Chrome
- Open Chrome and navigate to
chrome://extensions/ - Enable "Developer mode" (toggle in top-right corner)
- Click "Load unpacked" and select the
distfolder - The Seer extension should now appear in your extensions list
- Open Chrome and navigate to
Seer requires several components to function properly. Complete these steps in order:
- Install Anki if you haven't already
- Install the AnkiConnect add-on:
- In Anki: Tools → Add-ons → Get Add-ons
- Enter code:
2055492159 - Restart Anki
- Ensure Anki is running when using Seer. You can identify that it works by checking the Anki icon in the Options page after loading Seer.
-
Access Options Page
- Right-click the Seer extension icon → "Options"
- Or go to
chrome://extensions/→ Seer → "Extension options"
-
Download Frequency Data
- This extension requires JPDB data to be downloaded. This can be done from the options page manually, though the extension should do this automatically if you don't have it.
- This may take a few minutes and requires an internet connection
- Data is stored locally for offline use
-
Configure Vocabulary Sources
- Click "Add New Source" in the Vocabulary Sources section
- Source Name: Give it a descriptive name (e.g., "Core 2K")
- Deck Name: Select the deck you want to use for vocabulary tracking
- Field Name: Select the field containing Japanese words (usually "Expression" or "Word")
- Save the source
The stats section should update with the number of words in the deck. Multiple sources can be configured (for example, if you have Kaishi 1.5k AND a mining deck)
-
Set Up Ignored Words (Optional)
- Enable "ignored words functionality" if you want to mark words as ignored
- Configure deck name (default: "SeerIgnored")
- Set note type and field name
- Click "Setup Deck & Note Type" to create the necessary Anki components. You can also select your own.
-
Check AnkiConnect Status
- In the options page, look for the Anki status indicator
- Should show "Connected" with a green indicator
- If red, ensure Anki is running and AnkiConnect is installed
-
Test on a Japanese Website
- Visit any Japanese website (e.g., NHK News)
- Unknown words should be highlighted based on your vocabulary sources
- Click the extension icon to toggle highlights on/off
- Word Highlighting: Unknown words are automatically highlighted based on frequency and your Anki knowledge. Known or ignored words are not highlighted.
- Stats Panel: Click the floating stats button 🔮 to view reading statistics and quickly identify good pages for mining.
- Ignore Words: Alt+click any word to mark it as ignored
Seer offers multiple highlighting styles (configurable in options):
- Underline: Subtle underline highlighting
- Background: Background color highlighting
- Outline: Border outline highlighting
- Dots: Dotted underline highlighting
Words are color-coded by frequency:
- Very Common (1-1000): Light blue
- Common (1001-5000): Blue
- Uncommon (5001-15000): Orange
- Rare (15001-30000): Red
- Very Rare (30000+): Dark red
Enable this mode to highlight sentences containing exactly one unknown word:
- Ideal for comprehensible input
- Helps identify optimal learning material
- Toggle via popup or options page
- AnkiConnect integration for real-time deck synchronization
- JPDB frequency data for accurate word difficulty assessment
- Works with Yomitan for dictionary lookups and parsing
- Check AnkiConnect: Ensure Anki is running and AnkiConnect is installed
- Verify permissions: Extension needs access to all websites
- Reload extension: Go to
chrome://extensions/and reload Seer - Check console: Open DevTools (F12) and look for error messages
- Verify vocabulary sources: Check that deck names and field names are correct
- Test AnkiConnect: Use the "Check Connection" button in options
- Frequency data: Wait for JPDB data to download completely
- Field configuration: Ensure the field name matches your Anki deck structure
- Deck selection: Verify you're using the correct deck name
- Data sync: Try refreshing vocabulary sources in options
Enable debug mode in options for detailed logging:
- Go to Options → Debug section
- Enable "Debug Mode"
- Open browser console (F12) to see detailed logs
- Useful for diagnosing connection and parsing issues
If you encounter issues:
- Check the Issues page for known problems
- Enable debug mode and check console logs
- Create a new issue with:
- Browser version
- Extension version
- Console error messages
- Steps to reproduce
- TypeScript for type-safe JavaScript
- esbuild for fast compilation and bundling
- Chrome Extensions Manifest V3
- IndexedDB via Dexie for local data storage
- Kuromoji for Japanese text tokenization
- Yomitan: Pop-up dictionary for Japanese (highly recommended companion)
- AnkiConnect: Anki add-on for external integrations
- JPDB: Japanese word frequency database (data source)
- ASBPlayer: Video player for mining content