https://github.com/bigbrozer/linkwarden-aitiquette

Rust app for auto-labeling Linkwarden links using AI
https://github.com/bigbrozer/linkwarden-aitiquette

Last synced: 9 days ago
JSON representation

Rust app for auto-labeling Linkwarden links using AI

• Host: GitHub
• URL: https://github.com/bigbrozer/linkwarden-aitiquette
• Owner: bigbrozer
• License: mit
• Created: 2024-10-14T21:17:48.000Z (2 months ago)
• Default Branch: main
• Last Pushed: 2024-11-28T17:54:38.000Z (28 days ago)
• Last Synced: 2024-12-09T13:29:38.663Z (17 days ago)
• Language: Rust
• Homepage:
• Size: 39.1 KB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# Linkwarden Aitiquette

> The noun *aitiquette* is the combination of *ai* (artificial intelligence) and *étiquette* (the french word for *label*).

A Rust app for auto-labelling links in [Linkwarden](https://linkwarden.app/) using any [OpenAI](https://openai.com/) compatible APIs (works with [Ollama](https://github.com/ollama/ollama)).

*THIS IS CURRENTLY A WORK IN PROGRESS...*

## Implementation

To update tags for a link, here is the minimal request body:

```json

{

    "collection": {

        "id": 1,

        "ownerId": 1

    },

    "tags": [

        {

            "name": "test"

        }

    ]

}

```

## Next steps

This is a list of features I would like to implement:

- [ ] Add an option to perform tagging on non-tagged links only or all.

- [ ] Allow to select the language for tags.

- [ ] Add support for more AI models (e.g. GPT-3, BLOOM, etc.)

- [ ] Customize prompts.

- [ ] Temperature control for prompts.

- [ ] Do not load and store all links at once for large instances. Study Iterators.