The Agent Skills Directory

[COMMAND_EXECUTION]: The skill generates and instructs the execution of system-level commands to create symbolic links across different operating systems.
Evidence: workflows/create-symlink.md provides templates for ln -s (Linux/macOS), mklink /D (Windows CMD), and New-Item -ItemType SymbolicLink (PowerShell).
Risk: Commands use unvalidated placeholders for directory paths which could lead to command injection if paths contain shell metacharacters.
[COMMAND_EXECUTION]: The skill explicitly requires elevated permissions for specific operations on Windows environments.
Evidence: workflows/create-symlink.md states that mklink and PowerShell symbolic link creation require "Administrator privileges" or "Developer Mode". This encourages the agent to seek or use high-privilege execution environments.
[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection due to its processing of untrusted external content.
Ingestion points: workflows/create-prd.md accepts untrusted data from "initial prompt or attached files (txt, md, pdf)".
Boundary markers: Absent. The instructions do not define delimiters or warnings to ignore instructions embedded within the provided data.
Capability inventory: The skill has the capability to write files to the project directory (SKILL.md, workflows/create-prd.md) and execute system commands for directory linking (workflows/create-symlink.md).
Sanitization: Absent. There is no validation or escaping of the content extracted from external files before it is processed by the agent to generate documentation.

blueprint