The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/input_validation.py uses subprocess.run to call a project-local executor (.plan/execute-script.py) to retrieve a lesson inventory. This call uses a hardcoded list of arguments and does not invoke a shell, which prevents command injection risks.\n- [SAFE]: The skill implements centralized validation for multiple identifier types and file paths. The validate_relative_path function specifically detects and rejects absolute paths and directory traversal sequences, providing a robust defense for downstream scripts that handle file system operations.\n- [SAFE]: scripts/schema_validation.py provides field-level validation for project JSON storage files, enforcing type checks, allowed key sets, and maximum lengths to ensure data integrity.

tools-input-validation