README file from
GithubTag My Notes AI
Tag notes with your AI model of choice, choose which notes and which tags to consider and customize thought process to your preference.
Showcase
Quickly create an 'all-tags' operation on the active note

Tag a folder, active note, or all notes from the 'Tag My Notes' pop-up

Settings contains model & params, resoning steps and tag list

Features
- Choose which tags to apply, and which notes to apply them to
- See tagging progress live as the AI processes each tag for each note
- Tagging operations persist through closing and re-opening of the application
- View, cancel and delete tagging operations
- Customise the AIs train-of-thought to get more accurate results
- Supports the Vercel Gateway, OpenAI, Ollama, OpenRouter and Mistral with custom endpoints
Usage
In the ribbon on the left, select Tag My Notes to open the Tag My Notes pop-up.
From the Tag My Notes pop-up, you can choose which notes to tag and which tags to apply.
In the context menu (right-click on desktop), select Apply all tags to note to quicky tag the active note with all tags without having to open the Tag My Notes pop-up.
From the settings menu, add tags and their descriptions, choose an AI model, modify model parameters and even chain of thought.
Sponsor
Had to buy Vercel credit to test this so consider sponsoring if you find this useful.
Misc
[!IMPORTANT] This is a powerful tool. Adding more reasoning steps significantly increases token cost and tagging time. For processing a large number of files, it is best to use one reasoning step and to research the best model for your notes. Make sure to keep {description} and {tag} placeholders whenever you change reasoning steps.
[!NOTE] Network use disclaimer, this plugin requires a third party service unless you are self hosting the AI model. Vercel Gateway, OpenAI or Ollama are the available options.
Attribution
- Uses the JS AI API here
- Among libraries from the ai-sdk, the libraries ollama-ai-provider-v2 and @openrouter/ai-sdk-provider are used to add support for models.