Bucket Metadata Storage
Metadata storage is designed for structured key/value state.
Use it for lightweight persistent context that workflows can read and update.
Best Use Cases
- workflow run state markers
- threshold/config snapshots
- per-site operational context
Data Design Guidance
- use stable key names
- avoid deeply nested unbounded objects
- version metadata schema when changing structure
Workflow Pattern
ReadMetadata -> Decision/Transform -> SaveMetadata
This pattern supports stateful logic across executions.
Common Problems
- uncontrolled metadata sprawl
- hidden coupling from undocumented keys
- overwrite conflicts when multiple flows update same keys