The Agent Skills Directory

Indirect Prompt Injection (HIGH): The skill processes untrusted external content (Wardley Maps) and possesses file-modification capabilities.
Ingestion points: The skill reads and analyzes files using the Read and Grep tools.
Boundary markers: Absent. There are no instructions or delimiters defined to prevent the agent from following instructions embedded within the data files it analyzes.
Capability inventory: The skill is granted Write and Edit tool permissions, allowing for modification of the local filesystem.
Sanitization: Absent. There is no logic to sanitize or validate the content of the maps before processing.
Command Execution (MEDIUM): The skill requests broad filesystem access tools (Write, Edit) that exceed the requirements for a purely analytical strategic identification task, increasing the potential impact of an injection attack.
Data Exposure & Exfiltration (LOW): The skill instructions mandate external lookups via the 'perplexity' MCP server. This constitutes a network operation to a non-whitelisted domain, though it is used for information retrieval rather than exfiltration in this context.

strategic-plays