Skip to main content

​
What is AI Tag Generation?

AI tag generation uses artificial intelligence to analyze your bookmarks and automatically suggest relevant tags. This feature saves time and helps maintain consistent tagging across your collection. The AI considers:
  • The bookmark’s URL
  • The page title
  • The description metadata
  • Your existing tags from other bookmarks
The AI is designed to suggest tags that fit your existing tagging system, helping you maintain consistency and avoid creating duplicate or similar tags.

​
How It Works

When you trigger tag generation for a bookmark, the system:
1

Analyzes the bookmark

The AI examines the bookmark’s URL, title, and description to understand the content and context.
2

Reviews existing tags

It looks at all tags currently used across your bookmark collection to understand your organization system and preferred terminology.
3

Generates suggestions

Based on the analysis, the AI generates a set of relevant tags that match the bookmark’s content and fit your existing tagging conventions.
4

Applies tags automatically

The suggested tags are automatically applied to the bookmark - no manual review required.
The more bookmarks you have with well-organized tags, the better the AI becomes at suggesting tags that match your personal organization style.

​
Generating Tags for a Bookmark

1

Open bookmark options

Click the Options button on any bookmark card.
2

Select Generate tags

Click the Generate tags option from the dropdown menu. This option features a sparkle icon for easy identification.
3

Wait for processing

You’ll see a loading notification: “Crafting the perfect tags for you…” while the AI processes the bookmark.
4

Tags applied

Once complete, the tags will be automatically added to your bookmark and you’ll see a success notification.
The AI tag generation feature requires an active internet connection and authentication, as it makes a request to a secure Supabase Edge Function.

​
When to Use AI Tag Generation

​
Ideal Use Cases

New bookmarks - Generate tags immediately after adding a bookmark to keep your collection organized from the start. Bulk organization - Quickly tag multiple untagged bookmarks to bring your entire collection up to date. Consistency check - Use AI suggestions to ensure you’re using consistent tag names across similar content. Discovery - The AI might suggest tags you hadn’t thought of, helping you discover new ways to organize your content.

​
When to Tag Manually

Some situations where manual tagging might be better:
  • Personal projects - Bookmarks related to specific personal projects may need custom tags the AI won’t know about
  • Unique categories - If you have a very specific organization system, manual tagging gives you more control
  • Highly specific content - Niche or specialized bookmarks may need domain-specific tags
You can always manually edit AI-generated tags after they’re applied. Use AI for a quick start, then refine manually if needed.

​
Understanding the AI Process

​
Data Sent to the AI

The AI tag generation function receives: 
{
  bookmarkData: `${url} ${title} ${description}`,
  availableTags: ["design", "work", "tutorial", ...]
}
This combines all relevant text about the bookmark and provides context about your existing tagging system.

​
Available Tags Context

The AI receives a list of all unique tags currently used in your collection. This helps it:
  • Prefer existing tags over creating new ones
  • Match your naming conventions (e.g., “dev” vs “development”)
  • Understand your organization categories
  • Maintain consistency across your collection
The available tags are extracted from all bookmarks in your collection using a deduplication process that creates a unique set of all tags.

​
Handling Errors

If tag generation fails, you’ll receive an error notification:
“Failed to generate tags. Please try again later.”
Common reasons for failure:
  • Network issues - Check your internet connection
  • Authentication problems - Ensure you’re logged in
  • Server unavailable - The AI service may be temporarily down
  • Invalid bookmark data - The bookmark may be missing required metadata
If tag generation fails, the bookmark remains unchanged. Your existing tags (if any) are preserved.

​
Privacy and Security

The AI tag generation feature is designed with security in mind:
  • Authentication required - Requests include your session access token
  • Secure endpoint - Uses Supabase Edge Functions with built-in security
  • Data minimization - Only sends necessary bookmark data (URL, title, description)
  • No permanent storage - Bookmark data is processed in real-time and not stored by the AI service
Your bookmark URLs, titles, and descriptions are sent to the AI service for processing. If you have concerns about sensitive bookmarks, use manual tagging instead.

​
Best Practices

​
Start with Good Metadata

The AI relies on bookmark metadata for context. For best results:
  1. Ensure bookmarks have accurate titles (use Edit metadata if needed)
  2. Add descriptions when available
  3. Use Refresh metadata to get the latest information from websites

​
Build a Tag Foundation First

Before heavily relying on AI tag generation:
  1. Manually tag 10-20 bookmarks to establish your organization system
  2. Use consistent naming conventions
  3. Create tags that represent your main categories
This gives the AI a better foundation to work with.

​
Review and Refine

After AI generates tags:
  1. Review the suggested tags to ensure they’re appropriate
  2. Add any missing tags the AI didn’t include
  3. Remove tags that aren’t relevant
  4. Use the manual tag editing interface to make adjustments
The AI looks at your existing tags each time it runs. As you continue to organize your bookmarks and maintain consistent tagging, the AI’s suggestions will become more aligned with your preferences.Key actions that improve AI performance:
  • Maintain consistent tag naming across bookmarks
  • Reuse existing tags rather than creating similar ones
  • Keep your tag system organized and logical
  • Manually edit AI-generated tags when they’re not quite right

​
Combining AI with Manual Tagging

The most effective approach often combines both methods:
  1. AI first - Use AI tag generation to get a quick start
  2. Manual review - Check the suggested tags for accuracy
  3. Add specifics - Include project-specific or personal tags the AI wouldn’t know
  4. Remove irrelevant - Delete any tags that don’t fit
This hybrid approach gives you the speed of AI with the precision of manual curation.

Build docs developers (and LLMs) love

Get started for freeTalk to us