Serverless Cells
Serverless cells connect a URL to a notebook cell. When the URL is called, the latest notebook is fetched from the public internet and run in a private environment using headless browser. You can figure out what code will be run, in advance, based on the URL alone. Of course, because the source is defined within an Observable notebook also implies the source code can be edited and forked inside a web browser too. All taken together this approach has some nice properties. It's faster to redeploy code than
observablehq.com