The Agent Skills Directory

[COMMAND_EXECUTION]: The skill invokes a local script named memory-manager.py with various arguments to handle synchronization tasks, content diffing, and storage status updates.
[PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection (Category 8) by ingesting data from insights.json and the campaigns/ directory and storing it in persistent memory layers.
Ingestion points: The skill reads content from insights.json and files located in the campaigns/ directory.
Boundary markers: There are no explicit delimiters or instructions to the agent to ignore embedded commands within the ingested content.
Capability inventory: The skill utilizes memory-manager.py to interact with external vector databases and update local configuration files.
Sanitization: The process description does not specify sanitization or validation of the ingested content before it is stored in the memory layer.

sync-memory