For clay example client store build cache locally in the same directory as the scraper

[mlandauer]

This has the advantage of not needing any "special" caching directories locally rather we just have a simple naming convention that indicates a certain hidden file is always the build cache.

We would need to make the following changes:

• When the client uploads the build cache it would look for it in scraper_directory/.clay-build-cache.tgz.
• When the client downloads the latest build cache it would put it back in the same place
• When the client uploads the application code it shouldn't include the hidden file .clay-build-cache.tgz