Skip to content

kagisearch/kite-public

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Kite - News. Elevated.

MIT License CC BY-NC 4.0

This repository contains public files for Kite, news app for people who want a healthy news diet, developed by Kagi.

Kite is designed for people who want to stay informed without getting overwhelmed. We provide thrice-daily updates of the most important news stories, from carefully curated sources and summarized by advanced language models to give you the essential information you need.

Most of what powers Kite is open sourced and is found in this repository.

This includes

  • Kite web app
  • Community curated feeds
  • Media information

Core Principles

  • Updated only three times per day - no endless scrolling
  • Facts and perspectives, opinion-free
  • Zero tracking, zero ads
  • Pure signal, no noise
  • Quality over quantity
  • Complete news diet in 5 minutes

You can read more about core principles behind Kite

If you prefer to watch a short video, check this Ted talk: Four reasons you should stop watching the news | Rolf Dobelli.

Install & run Kite front-end

Kite front end is a statically served app and is fully open source.

Here is how to run it locally.

pip install Jinja2
python generate.py
open index.html

Kite front-end uses Kite application data that can be found at kite.kagi.com/kite.json (explore other files from there). Note that kite.json and files referenced by it are licensed under CC BY-NC license. This means that this data can be used free of charge (with attribution and for non-commercial use). If you would like to license this data for commercial use let us know through [email protected].

Kite web app is just one example front-end that one can run on top of the Kite data. We encourage others to contribute improvements to the Kite frontend.

We would also love to see what kind of custom front-ends you can create on top of Kite data! Feel free to share them with us and others by editing this Readme file.

Editing categories

To edit community curated categories, submit a pull request editing kite_feeds.json. If you do not know how to do that, you can open an issue and share the feeds you want to add there.

This file contains RSS feeds for various categories.

To add a new category or modify existing ones, follow this structure:

{
  "Category Name": {
    "feeds": [
      "https://example.com/rss-feed-1",
      "https://example.com/rss-feed-2"
      // ...
    ]
  },
  "Another Category": {
    "feeds": [
      "https://another-example.com/rss-feed-1",
      "https://another-example.com/rss-feed-2"
      // ...
    ]
  }
}

Adding a new category

  1. Create a new object with the category name as the key
  2. Add a "feeds" array containing the RSS feed URLs for that category
  3. Ensure proper JSON formatting (commas between objects, no trailing comma)

All categories are created equal and will appear as top level categories in the Kite app.

Ideas for categories:

  • Local news (city/state level)
  • Regional news (country/region)
  • Topical news (health, machine learning, aviation ...)

Guidelines for adding RSS feeds

Kite does not scrape websites, but only publicly available RSS feeds.

When adding an RSS feed make sure to:

  • Check that feed is working and has recent (daily) content.
  • Choose sources that have high quality content. Do not use low quality/gossip/SEO content.
  • Feeds should be in the same language as the rest of the feeds in that category (do not mix languages)

Important

We require at least 25 feeds for a category in order to surface it in Kite news. This is to make sure we maintain high level of quality of events covered in the app. The more high quality feeds exist for a category, the better Kite coverage will be. The feeds should be using the same language.

Kite does not scrape news websites, it only uses publicly available information in RSS feeds.

Guidelines for editing Media information

Kite uses contents of media_data.json to show additional information about sources of information. Initial information has been sources from https://statemediamonitor.com/ and the classification methodology is explained here https://statemediamonitor.com/methodology/

Feel free to add additional information (by editing media_data.json) both for privatly owned and state funded media organization. Add your sources of information in the pull request.