The Agent Skills Directory

[SAFE]: The skill functions as a reference for pandas-based data manipulation workflows. All code snippets and patterns align with standard data science practices and do not exhibit malicious intent.- [COMMAND_EXECUTION]: There is no evidence of shell command execution, subprocess spawning, or unauthorized system access. The scope is limited to Python library operations.- [DATA_EXFILTRATION]: No network operations (e.g., curl, wget, requests) or access to sensitive directories (e.g., .ssh, .aws) are present in the code or instructions.- [REMOTE_CODE_EXECUTION]: While the skill mentions dynamic tools like pd.eval() and numba.jit for performance optimization, these are used within the context of local DataFrame processing and do not incorporate untrusted external input.- [PROMPT_INJECTION]: The skill's instructions and constraints are focused on coding efficiency and data integrity, with no attempts to bypass safety filters or override agent behavioral constraints.

pandas-pro