By fast I mean two things: Files are hashed in parallel, so large sets can be processed quickly. On repeated runs, unchanged files skip hashing with a default

By fast I mean two things: Files are hashed in parallel, so large sets can be processed quickly. On repeated runs, unchanged files skip hashing with a default 90% probability using a cache. This keeps checks lightweight even at scale