The Agent Skills Directory

[COMMAND_EXECUTION]: The skill's workflow instructions involve running shell commands that include variables {sync_dir} and {prefix} derived from user input. This creates a potential surface for command injection if these inputs contain shell metacharacters.
[EXTERNAL_DOWNLOADS]: The skill downloads and installs the openpyxl and pyyaml libraries from the official Python Package Index (PyPI). These are standard dependencies for the skill's functionality.
[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it processes untrusted Drupal YAML configuration files and writes their contents into Excel cells without sanitizing for formula injection (CSV/Excel injection).
Ingestion points: Drupal configuration files (YAML) located in the user-specified sync directory (e.g., node.type..yml, taxonomy.vocabulary..yml).
Boundary markers: None present in the prompt interpolation or script logic to separate data from instructions.
Capability inventory: The skill performs file writing via openpyxl and executes subprocesses via the command line for dependency installation and script execution.
Sanitization: The script includes a strip_html function but lacks validation or escaping for Excel formula prefixes (e.g., =, +, -, or @) that may be present in the configuration data.

data-dict-extractor