🚀 The content of the web, accessible via a futuristic natural language interface.
JavaScript HTML CSS Shell
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Failed to load latest commit information.


Research (Work in Progress)

brought to you as libre software with joy and pride by Artificial Engineering.

Support our libre Bot Cloud via BTC 1CamMuvrFU1QAMebPoDsL3JrioVDoxezY2.


Project Research is a tool that helps you understand the internet more easily and comes with an easy natural language interface. It is not a stupid assistant, but more of an "internet buddy" to ease up your web browsing experience and support you as far as possible.



  • does not load Cookies, JS or CSS
  • does not load Flash, Advertisements, iFrames or trackable resources
  • totally anonymous, saves much bandwidth and re-uses offline cache
  • sweet NLP interface that eases up indexing, crawling and scraping tasks
  • reproducible history, allowing forward/backward modifications in time
  • cache uses markdown to represent results (and saves storage space)

You have full control over what content is requested. There's even a Tethering Mode that will allow you to only load text content and will not load images or videos until you say so.

Project Research also saves you bandwidth. By default, it will prefer the local remembered content and not request anything until you say so. You have full control about what content is loaded.

It can archive ANY kind of content to a simple shareable offline archive folder on your hard drive that you can share via USB and reuse with other Project Research installations on other computers.

It is peer-to-peer and allows sharing your offline archive of websites with other local instances, so that you don't waste bandwidth on a shared internet connection.

Under the hood it uses nw.js and lychee.js and has a zero-unnecessary-dependency philosophy.


For now, we don't have ready-to-use binaries. This project is still in alpha, but soon we may have some ready for you in the releases section.

Research is a project made with lychee.js.

It is automatically built and deployed to GitHub using the following lycheejs-fertilizer integration scripts:

  • bin/build.sh builds the project
  • bin/package.sh bundles the project into binaries (nwjs platform)
  • bin/publish.sh publishes the binaries to a new GitHub Release.

Work-in-Progress (aka not working)

These are the features that are currently work-in-progress and are known to fail right now. If you want to help building these, you are welcome to submit an Issue or a Pull Request. We could use any help :)

  • Fallback Reader Functionality
  • Imgur Plugin
  • Facebook Plugin
  • Gfycat Plugin
  • Instagram Plugin
  • Medium Plugin
  • DuckDuckGo Plugin

Usage / How To Test (for now)

Builds are released once we get into public beta and have above plugins implemented. For now, you manually have to download the nw.js SDK and execute it in Project Research's root folder:

  • Have a UNIX machine (Linux, OSX, BSD, whatever)
  • Install nw.js SDK on your machine
  • Do this in your Terminal:
git clone https://github.com/Artificial-Engineering/research.git;

cd ./research;
nw .;


  • @<person> for person related search
  • #<topic> for topic related search
  • ~<location> for location related search
  • <input> for generic search

Incremental Schema

  • sort by relevance
  • sort by votes or sort by likes
  • sort by time or sort by date


  • app.interface.Intent needs time support (from x to y? from the last x weeks/days/months?)

  • app.plugin.Reddit needs time support

  • Figure out how to influence Rendering

  • Browse State > Reddit Plugin > Reddit Result?

  • Maybe some article class="reddit" or so?

  • Better architecture would be using Polyfillr Components, so <reddit-article> would just inherit correctly


Project Research is released under GNU GPL 3 license.