The Agent Skills Directory

[SAFE]: A comprehensive security audit of all files, including SKILL.md and referenced documentation, revealed no malicious patterns. The instructions align with the stated purpose of providing senior-level data engineering guidance.
[SAFE]: The Python scripts in the scripts/ directory (pipeline_orchestrator.py, data_quality_validator.py, and etl_performance_optimizer.py) are boilerplate CLI skeletons. They do not import third-party libraries, perform network operations, or execute system commands. They follow standard practices for logging and argument parsing.
[INDIRECT_PROMPT_INJECTION]: The scripts define a potential ingestion surface for untrusted data by accepting an --input path, though no active processing is performed.
Ingestion points: The main() function in all scripts located in the scripts/ directory accepts input via the --input command-line argument.
Boundary markers: No specific boundary markers or instructions to ignore embedded instructions are present in the scripts or documentation.
Capability inventory: No dangerous capabilities such as file system writes, network requests, or shell command execution are present in the provided source code.
Sanitization: No input sanitization or validation logic is implemented beyond basic argument parsing, as the scripts are currently non-functional templates.

senior-data-engineer