The Agent Skills Directory

[Indirect Prompt Injection] (LOW): The skill creates a surface for malicious instructions to be embedded in data processed at runtime.
Ingestion points: User-provided inputs including 'App name', 'Homepage', and 'Download URL' collected in SKILL.md.
Boundary markers: Absent; the workflow interpolates these values directly into a Ruby code template in references/homebrew-cask-contribution-workflow.md.
Capability inventory: The generated Ruby file is executed on the host system via brew install as described in workflow.md.
Sanitization: None; there is no validation or escaping of the user-provided strings before they are written to the script file.
[Dynamic Execution] (LOW): The skill involves script generation and execution by assembling Ruby .rb files at runtime based on user templates and subsequently running them.
Evidence: references/homebrew-cask-contribution-workflow.md details writing the cask definition and executing it via brew install.
Context: This behavior is inherent to the primary purpose of authoring Homebrew Casks.
[Command Execution] (SAFE): The skill utilizes standard developer commands and filesystem operations consistent with its stated purpose.
Evidence: Uses brew style, brew audit, brew install, and ln -s to manage local Homebrew taps.

homebrew-cask-authoring