Checks if an entity exists in persistent storage by its ID. Useful for conditional operations without reading the full entity (e.g., checking session memory existence).
Reads an entity from persistent storage by its ID, parsing it from a JSON file. Used to retrieve persisted data such as agent states, memory, or alive status.
Initializes the storage directory, creating it if needed and validating existing data by removing invalid entities. Ensures the persistence layer is ready for use, handling corrupt files during setup.
Writes an entity to persistent storage with the specified ID, serializing it to JSON. Uses atomic writes to ensure data integrity, critical for reliable state persistence across swarm operations.
Defines the core interface for persistent storage operations in the swarm system. Provides methods for managing entities stored as JSON files in the file system, used across swarm utilities.