run-python-safely

To safely execute agent-generated Python code, this skill applies AST-based analysis to block dangerous imports, builtins, and filesystem-modifying methods before running inline code or scripts.

Yes, you can run Python files from automated agent workflows by passing the file path to the execution script, which analyzes the AST and blocks dangerous operations prior to execution.

AST-based analysis parses Python code into an abstract syntax tree to detect and prevent dangerous imports, builtins, and filesystem-modifying methods from executing in automated pipelines.

To set up a timeout for dynamically generated Python scripts, this skill enforces a strict 5-minute execution limit and returns standardized exit codes for timed-out or blocked runs.

Yes, this Python execution skill works with inline code strings by accepting a command-line argument, allowing you to quickly test and execute dynamically generated snippets safely.

A limitation of using AST analysis for Python code safety is that it specifically blocks dangerous imports, builtins, and filesystem-modifying methods, while enforcing a 5-minute timeout to prevent runaway processes.