The Agent Skills Directory

[DATA_EXFILTRATION]: The shopify_init.py script identifies and reads sensitive environment files to load configuration variables. It searches for .env files in the current working directory and its parents, specifically including the .claude configuration directory. Accessing environment files outside the immediate project scope is a security concern as it may expose unintended credentials from the host environment.
[COMMAND_EXECUTION]: The shopify_init.py script utilizes subprocess.run to execute the shopify version command locally. This is used as a functional check to ensure the necessary CLI tools are installed.
[EXTERNAL_DOWNLOADS]: The skill documentation and initialization script refer to the installation and use of official Shopify CLI tools and npm packages. These resources originate from a well-known service (Shopify) and are documented neutrally as intended dependencies.
[PROMPT_INJECTION]: The shopify_init.py script collects user input through the input() function and interpolates it into configuration templates (e.g., shopify.app.toml) using f-strings without validation or escaping. This creates a vulnerability to indirect prompt injection or schema confusion if the generated files are subsequently read by an automated agent.
Ingestion points: input() calls in the ShopifyInitializer.prompt and ShopifyInitializer.select_option methods in scripts/shopify_init.py.
Boundary markers: None; user input is directly placed into double-quoted string fields in TOML and JSON templates.
Capability inventory: The skill has the capability to write files (Path.write_text) and execute subprocesses (subprocess.run) as seen in scripts/shopify_init.py.
Sanitization: No sanitization or validation of the input content is performed beyond a trailing whitespace strip() call.

shopify