Automatic short-form explainer videos for the news
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Type Name Latest commit message Commit time
Failed to load latest commit information.
requirements.txt lets you semi-automatically generate a background explainer video for any story, inspired by "Previously On..." recap sequences on TV shows. It uses a story's inline links to generate a network of possible stories; after selecting the headlines and images that matter most, it creates a video of them using Stupeflix.



  • git clone this repo and cd into it, activate virtualenv if you like
  • pip install -r requirements.txt installs everything
  • set the STUPEFLIX_API_KEY and STUPEFLIX_API_SECRET environment variables to your Stupeflix keys (or change them in, but don't re-commit!)
  • python starts the server
  • navigate to localhost:5000

Adding Backstories

  1. go to and find the DBS dictionary
  2. add an item to the dictionary where the key is the name of the database you want to create, and the value is the URL to start crawling from, e.g. "'ukraine': ''"
  3. to start the crawl, run python <DB_NAME>
  4. to access the results, find them in the dropdown in localhost:5000


  • / - leads to homepage
  • /stupefy.json - takes a list of stories in POST data and generates video via Stupeflix API
  • /stupestatus.json - checks status of a video given its Stupeflix key


  • templates/index.html contains all the frontend code
  • contains the server and endpoints for talking to the frontend
  • extracts and indexes links from articles
  • contains the crawler and higher-level interaction with the database
  • is for lower-level integration with MongoDB (stolen from MediaCloud-API-Client)
  • integrates with the Stupeflix API