The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute a provided Python script (create_storyboard_package.py) which initializes the production environment by creating directories and writing files. The script uses user-supplied arguments, such as the project slug, to construct file paths (e.g., root / "storyboard_projects" / args.slug) using pathlib without sanitization. This logic permits path traversal, potentially allowing the agent or an attacker to create directories or files outside the intended workspace if the agent does not validate the input.
[PROMPT_INJECTION]: The skill processes untrusted user-provided content, creating a vulnerability to indirect prompt injection.
Ingestion points: The agent is directed to "Extract the brief" and "Analyze the script" from user-provided inputs as part of its standard workflow.
Capability inventory: The agent possesses the capability to execute local shell commands and perform arbitrary file system writes through the provided scaffold script.
Boundary markers: The skill lacks explicit boundary markers or instructions to the agent to disregard instructions embedded in the processed user data.
Sanitization: No sanitization or validation steps are defined to filter malicious instructions from the ingested script or brief content.

create-storyboard