/mstack:idea-shotgun

Stage: ideate · Voice: generative

Produces variants so the user picks from a set instead of a hunch.

Procedure

1. Load .mstack/research-question.md (if a prior attempt exists) and any data documentation.
2. Get the seed. One paragraph from the user: what they have (data, theoretical interest), what the question might be, the audience they imagine. If the conversation or .mstack/research-question.md already supplies these, do not re-ask.
3. Generate 4–6 angles, each varying along one axis:
   • same data, different question;
   • same question, different identification;
   • same identification, different scope (population, time, treatment);
   • same headline, different audience (IPE / comparative / methods / public).

4. Card each angle:

   ANGLE N: <one-sentence claim>
   Contribution:    <the new sentence this adds to the literature>
   Identification:  <the design, the quantity it identifies, and how it separates the effect from the obvious confound>
   Cost:            <data-acquire weeks | analysis weeks | total months>
   Risk:            <highest single risk: data, identification, scoop, fit>

5. Rank qualitatively on (contribution × tractability) / (cost × risk); surface the top 2.
6. Save every angle, including the unsurfaced ones (next year's paper), to .mstack/idea-shotgun-<YYYY-MM-DD>.md.

Outputs

• .mstack/idea-shotgun-<date>.md — full angle set with cards.
• Summary block: top 2 with a one-line case each, and which to take to /mstack:research-question.

Anti-patterns

• Variations on a theme. Five angles that swap one variable are not divergence.
• Defaulting to the seed. At least one angle should reframe the question entirely.

Next

/mstack:research-question on the chosen angle.