The Agent Skills Directory

[Indirect Prompt Injection] (HIGH): The skill processes untrusted external data (component source code) and possesses high-privilege capabilities including Bash execution and file Write access.
Ingestion points: Source code of components being tested (SKILL.md).
Capability inventory: Bash, Write, Edit, Read, Grep, Glob (SKILL.md).
Boundary markers: Absent. There are no instructions for the agent to ignore instructions embedded within the component code.
Sanitization: Absent. No validation or filtering of input code before processing.
Risk: A malicious component file could contain instructions designed to hijack the agent's logic to execute arbitrary shell commands or write malicious scripts to the filesystem.
[Command Execution] (MEDIUM): The skill explicitly allows the Bash tool. While necessary for some development tasks, its inclusion in a skill that processes user-provided code increases the risk of command injection if the agent is not strictly constrained.

frontend-test