The Agent Skills Directory

[Indirect Prompt Injection] (MEDIUM): The skill is designed to analyze and modify external React component files, which are untrusted data sources.
Ingestion points: Component source code accessed via Read, Grep, and Glob tools.
Boundary markers: The skill lacks explicit instructions to the agent to treat code comments or metadata as non-instructional, increasing the risk of the agent obeying embedded commands.
Capability inventory: The skill possesses Write and Edit permissions, allowing it to commit changes to the local filesystem based on processed input.
Sanitization: There is no evidence of content sanitization or validation before processing external code.
[COMMAND_EXECUTION] (LOW): While the skill uses standard tools like Grep and Glob, these are used within the context of code analysis and do not currently point to arbitrary shell execution.

validation-helper