The Agent Skills Directory

[COMMAND_EXECUTION]: The scripts/script_tester.py utility utilizes subprocess.run() to execute Python scripts discovered within a user-provided directory path.
[DYNAMIC_EXECUTION]: The skill performs runtime execution of local Python files as part of its testing framework. This allows for the execution of arbitrary code contained within a user-specified directory.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from external skill directories and possesses significant execution capabilities.
Ingestion points: The skill_path parameter accepted by skill_validator.py, script_tester.py, and quality_scorer.py.
Boundary markers: None present; the tools process all files within the provided directory.
Capability inventory: subprocess.run() in scripts/script_tester.py for executing scripts and checking help output.
Sanitization: The tools use ast.parse() to check syntax and imports, but do not prevent the execution of scripts that utilize dangerous standard library modules if the syntax is valid.

skill-tester