The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses local shell commands, including cp to initialize settings files and ruby to check for library existence and perform maintenance tasks.
[DYNAMIC_EXECUTION]: Step 3 utilizes ruby -e to execute a script snippet constructed from user-provided inputs like library names, languages, and media paths, which is a dynamic execution pattern.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user data (media paths and library names) that are interpolated into shell commands.
Ingestion points: User input for library names and media file paths collected during the interactive questions in Step 2.
Boundary markers (absent): The instructions provide normalization for library names but do not specify explicit delimiters or shell escaping for media paths when building the ruby -e command.
Capability inventory: The skill has capabilities to copy files (cp) and execute local Ruby scripts.
Sanitization (present): Library names are normalized to alphanumeric characters and dashes, and the agent is instructed to verify the existence of all media files before proceeding, which serves as a basic validation gate.

create-library