The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the Bash tool to execute file management commands (mv, mkdir) and developer tools via npx (e.g., npx nx run <app>:typecheck). These operations provide extensive control over the workspace.
[DYNAMIC_EXECUTION]: The skill assembles shell commands dynamically using file paths as arguments. Specifically, in step 4.2, it uses mv with paths obtained from a filesystem scan. This introduces a risk of command injection if filenames contain shell control characters (e.g., ;, &, |).
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data (filenames and file contents) to determine its plan of action. This attack surface could be used to manipulate the agent's logic through carefully crafted file structures or source code.
Ingestion points: Files and directory names within the user-provided TargetDirectory (SKILL.md).
Boundary markers: None identified for distinguishing between instructions and scanned data.
Capability inventory: mv, mkdir (Bash), Edit (file write), Read, Grep, Glob across multiple scripts.
Sanitization: The skill does not describe sanitization or escaping of file paths before they are used as shell command arguments.

enforce-standards