A resource is an open concept on the The Eye. If you head to the web UI they are the first thing you'll see, they are whatever can report a status and for that, as this concept, they deserve the main view of the UI dashboard. Basically every resource is a monitor watching things out for you.
Take a host for example: its status is unknown and unmeasurable. Only when a host monitor is assigned to it, and starts reporting it's status and health, the host becomes a reliable, accountable resource.
A web scrapper is a way of monitoring an external websites or APIs, can report a status and, if needed, trigger tasks or workflows. It's an accountable resource signaling monitor events.
As you can see, the resource concept points to be any source of information (thus, decision) on your infrastructure. Complex and dependent processes can be setup though resources and, as well as it may sound confusing first, it becomes more and more as a comfortable concept for it's defined functionality and schema.