Smart Tagger
approvedby Johan Denoyer
Automatically generate 2-5 relevant tags for your notes using ChatGPT or Claude AI. - This plugin has not been manually reviewed by Obsidian staff.
Smart Tagger for Obsidian
An Obsidian plugin that automatically generates 2-5 relevant tags for your notes using AI (ChatGPT or Claude).
Features
- Dual AI Support: Choose between OpenAI (ChatGPT) or Claude API
- Smart Tag Generation: Generates 2-5 contextually relevant tags per note
- Vault Tag Matching: AI-generated tags are matched against your existing vault tags — vault casing always wins
- Batch Processing: Process all notes in your vault at once
- Tag Preview: Review and select tags before applying them
- Customizable: Configure prompts, tag limits, and behavior
- Existing Tag Awareness: Optionally exclude tags that already exist
- Rate Limiting: Configurable requests-per-minute to avoid quota exhaustion
- Retry Logic: Automatic exponential backoff on transient failures
- Multi-language: English and French interface
Installation
Manual Installation
- Download the latest release from the GitHub releases page
- Extract the files to your vault's
.obsidian/plugins/smart-tagger/directory - Enable the plugin in Obsidian's Community Plugins settings
Building from Source
- Clone this repository
- Run
npm installto install dependencies - Run
npm run buildto build the plugin - Copy
main.js,manifest.json, andstyles.cssto your vault's.obsidian/plugins/smart-tagger/directory
Setup
- Go to Settings → Community Plugins → Smart Tagger
- Choose your preferred AI provider (OpenAI or Claude)
- Add your API key:
- OpenAI: Get your API key from OpenAI's platform
- Claude: Get your API key from Anthropic's console
- Configure other settings as needed
Usage
Generate Tags for Current Note
- Click the tag icon in the ribbon bar
- Use the command palette: "Generate tags for current note"
- The plugin will analyze your note content and suggest relevant tags
Batch Processing
- Use the command palette: "Generate tags for all notes"
- This will process all markdown files in your vault
- ⚠️ Warning: This may consume significant API credits for large vaults
- ⚠️ Warning: If you have not allowed "Auto Apply Tags" in settings, you will need to approuve tag additions on every note
Tag Preview
- By default, generated tags are shown in a preview modal
- You can select which tags to apply to your note
- Enable "Auto Apply Tags" in settings to skip the preview
Settings
- Default AI Provider: Choose between OpenAI and Claude
- API Keys: Configure your OpenAI and/or Claude API keys
- Tag Limits: Set minimum (1-10) and maximum (1-10) number of tags
- Custom Prompt: Customize the AI prompt for tag generation
- Auto Apply Tags: Skip preview and automatically apply generated tags
- Exclude Existing Tags: Don't generate tags that already exist in the note
- Prefer Existing Vault Tags: Match AI output against vault tags case-insensitively; vault casing wins (default: enabled)
- Rate Limit: Max API requests per minute, 1-60 (default: 10)
- Max Retries: Retry attempts on failure, 1-10 (default: 3)
- Language: Interface language — English or French
API Usage and Costs
This plugin makes API calls to external AI services:
- OpenAI: Uses GPT-3.5-turbo model (~$0.002 per 1K tokens)
- Claude: Uses Claude-3-haiku model (~$0.00025 per 1K tokens)
Each note processed typically uses 50-200 tokens depending on content length.
Privacy
- Your note content is sent to the selected AI provider for tag generation
- No data is stored by this plugin beyond your settings
- API keys are stored locally in your Obsidian vault
- Review the privacy policies of your chosen AI provider
Support
If you encounter issues or have feature requests:
- Check the existing GitHub issues
- Create a new issue with details about your problem
- Include your Obsidian version and plugin version
Development
This plugin is built with TypeScript and uses:
- Obsidian API for vault interaction
- OpenAI API for ChatGPT integration
- Anthropic API for Claude integration
- esbuild for bundling
License
MIT License - see LICENSE file for details
For plugin developers
Search results and similarity scores are powered by semantic analysis of your plugin's README. If your plugin isn't appearing for searches you'd expect, try updating your README to clearly describe your plugin's purpose, features, and use cases.