AIASys Notebook-first Workflow

To ensure reproducible data analysis, you should adopt a notebook-first workflow that splits code into logical cells and executes scripts directly within notebooks. This approach preserves context, prevents lost results, and eliminates the risks of ad-hoc REPL sessions.

The best way to manage large DataFrame outputs is to apply concise data observation best practices within your notebook workflow. This prevents large outputs from consuming excessive context, ensuring your analysis sessions remain fast and efficient.

You structure a notebook for data analysis by listing existing notebooks, creating scratch notebooks for experiments, and splitting your analysis into logical cells. Following explicit cell editing and execution rules prevents common errors like overwriting entire notebooks.

You can technically use ad-hoc REPL workflows, but they are not recommended because they cause lost context and unreproducible results. A notebook-first workflow is enforced to provide standardized management, cell editing, and execution for AIASys data analyst sub-agents.

To avoid overwriting entire notebooks during cell editing, follow the 12 targeted rules for notebook management that separate notebook and plain file editing semantics. These rules ensure safe execution and proper cell management for your analysis tasks.