The Agent Skills Directory

[Indirect Prompt Injection] (HIGH): The skill possesses a significant attack surface by combining the ingestion of untrusted data with file-modification capabilities.
Ingestion points: Processes external Product Backlog Items (PBIs) from the team-artifacts/pbis/ directory using the Read tool.
Boundary markers: There are no defined delimiters or instructions provided to the agent to disregard instructions potentially embedded within the PBI content.
Capability inventory: The skill is granted Write, Edit, and TodoWrite tools, allowing it to modify the filesystem and create tasks based on the data it processes.
Sanitization: No sanitization, validation, or escaping of the external content is performed before the agent uses it to make decisions or write updates.

prioritize