Why should CloudMD files not be excessively large?	The longer the file, the less reliably Cloud can follow the instructions.
Where should hard rules, such as "never push to main," be placed instead of CloudMD?	In a pre-tool use hook.
List the four locations where CloudMD instructions can exist.	Managed policy, user file, project file, and local.
What is the purpose of the "local" scope in CloudMD?	To hold personal architectural decisions specific to a project, such as refactoring.
What is the function of using path-to-file syntax (imports) in CloudMD?	To split up and organize large instructions into smaller, more maintainable files.
Do imports reduce the context load for CloudMD?	No, Cloud expands imported files in line at launch, so they do not reduce context.
What is the primary requirement for effective CloudMD rules?	They must be specific and checkable.
Why should named exports be used instead of default exports in CloudMD?	To close potential misinterpretations.
How should emphasis be used when writing CloudMD rules?	It should be spent only on the two or three rules that absolutely must be followed.
How should a user treat their CloudMD file for maintenance?	Treat it like production code; justify every line or delete it.