ds-brainstorming

What problem does it solve?

It helps when a notebook-based analytics request is still vague, so you can lock down the hypothesis, analysis unit, metric hierarchy, time window, and robustness checks before any execution happens.

Core Features & Use Cases

• Structured design first: Produces an explicit analytical design document rather than jumping into SQL/pandas/statistics.
• Clarifying questions and tradeoffs: Asks one question at a time, then proposes 2–3 approaches with risks and decision rationale.
• Safe handoff to implementation skills: Forces approval of the design and then transitions the work to ds-analysis-plan or ds-experiment-design.

Example: You’re asked to “check whether feature X affects conversion Y,” but you don’t yet know the appropriate randomization/analysis unit, which metrics should be primary vs guardrails, or which time windows and exclusions matter.

Quick Start

Use ds-brainstorming to define the analytical design for my question “Does X affect Y?” including unit, metrics, time window, and robustness checks, and only proceed after you get my approval.