Ask HN: Run discovery of hypermedia APIs

In terms of the discovery of hypermedia links in RESTful HATEOS APIs, can anybody provide any examples or information on how to implement such a dynamic traversal of documents in order for the client to 'discover' where to go next.

Of-course it is easy for the developer to hardcode which parts of the JSON response is important in order to decipher where it should go next, but in order to keep that loose coupling, whats the best way to approach this?

It seems examples and documentation on this subject is a little sparse.