The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill possesses a capability-rich attack surface because it is designed to ingest codebase data and execute actions based on that data.
Ingestion points: The skill uses Read, Grep, and Glob tools to explore local repository files (SKILL.md, SKILL.ko.md).
Boundary markers: The instructions do not specify explicit boundary markers or sanitization requirements for content read from the repository.
Capability inventory: The skill has significant capabilities, including the ability to write/edit files and execute arbitrary shell commands via Bash for validation and implementation.
Sanitization: There is no explicit mention of sanitizing or escaping content before it is processed or used in reasoning steps.
Risk: A malicious file within the repository could contain hidden instructions that influence the agent to perform unauthorized actions during the implementation or validation phases.

execute