Skip to main content
This guide walks you through getting your first tag suggestions with Auto Tagger. By the end, you’ll understand how to trigger suggestions, accept them, and configure the plugin for your workflow.

​
Prerequisites

Before starting, make sure you have:
  • Installed Auto Tagger in your Obsidian vault (see Installation)
  • At least a few notes with existing tags (the plugin needs 2+ documents per tag to learn patterns)
Auto Tagger learns from your existing tagging patterns. If your vault has no tags yet, manually tag 5-10 notes across different topics before proceeding.

​
Initial Vault Scan

When you first enable Auto Tagger, it automatically scans your vault to build its tagging model.
1

Wait for the scan to complete

After enabling the plugin, you’ll see a notice: “Auto Tagger: Scanning vault…”The scan completes when you see: “Auto Tagger: X tags learned from Y notes”
2

Check the scan results

Open Settings → Auto Tagger to see your model statistics:
  • Number of unique tags learned
  • Number of tagged documents
  • Total documents scanned
  • Unique words indexed
If you see “0 tags learned”, your vault doesn’t have enough tagged documents. Manually add tags to at least 2-3 notes, then use the Rescan vault button in settings.

​
Getting Tag Suggestions

Auto Tagger offers three ways to get tag suggestions:

​
1. Automatic Suggestions While Typing

The plugin automatically suggests tags as you type in any note.
1

Open or create a note

Open an existing note or create a new one with some content.
2

Write some content

Type at least a few sentences. The plugin needs enough text to analyze.
3

Wait for suggestions

After 2 seconds of inactivity (the default debounce delay), you’ll see a suggestion notice in the bottom-right corner:Suggest: #your-tag (+2 more)
4

Accept or skip

  • Click Add to insert the tag
  • Click Skip to ignore this suggestion
  • Click Skip all to dismiss all pending suggestions
  • Click Show all to see all suggestions in a modal

​
2. Manual Command Trigger

You can manually trigger suggestions at any time.
1

Open the command palette

Press Ctrl/Cmd + P to open the command palette.
2

Run the suggest command

Type “Suggest tags” and select “Suggest tags for current note”.
3

Review all suggestions

A modal appears showing all suggestions with confidence bars. Select the tags you want and click Add selected.

​
3. On File Open

When you open a note, Auto Tagger automatically checks if it should suggest tags.
  • If the note has no tags, you’ll get suggestions after a 500ms delay
  • If the note has some tags, it will suggest additional related tags

​
Accepting Tags

When you accept a tag, Auto Tagger inserts it according to your Tag placement setting:
Tags are inserted on the first content line (after frontmatter if present):
---
title: My Note
---

#productivity #workflow

Your note content here...
If the first line already has tags, the new tag is appended to that line.

​
Configuring Your Workflow

Fine-tune Auto Tagger for your needs in Settings → Auto Tagger:
Change the Check delay setting to control when suggestions appear:
  • Lower values (500-1000ms): More responsive, but may interrupt your writing
  • Higher values (2000-5000ms): Less intrusive, but slower to respond
Default: 2000ms
Set Max suggestions to limit how many tags are offered at once:
  • Lower values (1-3): Only show the most confident matches
  • Higher values (5-10): See more options, including less certain matches
Default: 5
Adjust Minimum confidence to control suggestion quality:
  • Higher values (0.02-0.05): Only highly relevant tags
  • Lower values (0.005-0.01): More suggestions, potentially less accurate
Default: 0.01
Turn off Auto-suggest while editing to only get suggestions via manual command or on file open.

​
Rescanning Your Vault

As you add more tags to your notes, you should rescan your vault to improve suggestions:
1

Open settings

Go to Settings → Auto Tagger.
2

Click Rescan vault

Click the Rescan now button in the Rescan vault section.
3

Wait for completion

A notice will show “Done: X tags from Y notes” when finished.
Rescan after:
  • Bulk-tagging multiple notes
  • Creating new tags
  • Reorganizing your tagging system

​
Tips for Better Suggestions

Tag consistently

Use the same tags for similar content. Inconsistent tagging confuses the model.

Use descriptive tags

Specific tags like #machine-learning work better than vague ones like #tech.

Tag multiple notes

Each tag needs at least 2 documents to generate suggestions.

Write substantial content

Longer notes with more distinctive words produce better suggestions.

​
Troubleshooting

Possible causes:
  • Your vault has fewer than 2 documents with the same tag
  • The minimum confidence threshold is too high
  • The note doesn’t have enough text to analyze
Solutions:
  • Add tags to more notes and rescan
  • Lower the minimum confidence in settings
  • Write at least 2-3 paragraphs of content
Possible causes:
  • Your existing tags are inconsistently applied
  • The minimum confidence threshold is too low
  • The vault hasn’t been rescanned recently
Solutions:
  • Review and standardize your existing tags
  • Increase the minimum confidence in settings
  • Click Rescan now in settings
Adjust the Check delay setting in Settings → Auto Tagger. Lower values respond faster but may interrupt your writing.

​
Next Steps

Learn How It Works

Understand the TF-IDF algorithm behind Auto Tagger

Explore Settings

Deep dive into all configuration options

Usage Tips

Best practices for optimal suggestions

Troubleshooting

Solve common issues

Build docs developers (and LLMs) love

Get started for freeTalk to us