The Agent Skills Directory

[COMMAND_EXECUTION] (SAFE): The scripts scripts/code_quality_analyzer.py, scripts/fullstack_scaffolder.py, and scripts/project_scaffolder.py use argparse to accept a target path but do not execute any shell commands or subprocesses. They are currently logic-free skeletons that only validate the existence of a path.
[DATA_EXFILTRATION] (SAFE): No network-related modules (like requests or urllib) are imported or used. There is no evidence of data being transmitted to external endpoints.
[INDIRECT_PROMPT_INJECTION] (LOW): The scripts take a target path as an argument. While this is a standard pattern for analysis tools, it represents a surface where a malicious project directory could contain filenames or content intended to confuse an agent processing the tool's output. However, the current scripts do not read or output file content, neutralizing this risk.
Ingestion points: target_path argument in all three scripts.
Boundary markers: Absent; output is plain text report summaries.
Capability inventory: Path validation (os.path/pathlib).
Sanitization: None; paths are validated for existence but not for content content.

senior-fullstack