The Agent Skills Directory

[SAFE]: The skill utilizes well-known, established libraries from the Python data science ecosystem such as Pandas, Matplotlib, and Scikit-learn. All external tool references and installation commands target legitimate, widely-used software from official registries.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process external datasets, identified as ingestion points in references/large-dataset-eda.md and references/profiling-automation.md. While explicit boundary markers and sanitization logic are not defined, the skill mitigates risk by emphasizing automated profiling and domain-knowledge validation as part of the core workflow. Capability inventory includes local server execution for visualization and file writing for saving artifacts.
[DYNAMIC_EXECUTION]: The skill documents the use of tools like Papermill for parameterized notebook execution and Bokeh/Streamlit for interactive applications, which involve standard dynamic execution patterns for data science workflows.
[COMMAND_EXECUTION]: Instructions include standard shell commands for starting development servers (e.g., bokeh serve, voila, and streamlit), which are essential for the skill's documented visualization tasks.

analyzing-data