Run a brainstorming session with AI
Explore the team’s own ideas before adding alternatives, then choose a small experiment with an explicit reason for trying it.
Run a brainstorming session with AI when you want more than a list of suggestions from the fastest voice in the room. The brainstorm-coach skill gives participants the first turn, then adds a few clearly labelled alternatives. A group trying to make delivery times easier to understand might start with the order screen, the receipt or the collection point.
Keep evaluation for a later pass. Once the options are visible, two related ideas can be grouped without losing who proposed them. An explicit reason for a choice is easier to challenge than an unexplained numerical score, and the next step can be small enough to try this week.
What it does
- Define a practical question and the constraints around it.
- Keep human suggestions and agent additions identifiable.
- Try a different angle when a round stops producing useful ideas.
- Group related proposals and state the reason for a preference.
- Choose a first experiment and the observation that would justify continuing.
Install and use
- Give a conversational agent the problem, intended outcome and authorised session notes. A live external collaboration account is unnecessary for a planning session.
- Ask, “Take our suggestions for explaining delivery times first. Then offer two distinct approaches and leave ranking until the next round.”
- Choose your own shortlist before comparing it with the agent’s suggestions.
FAQ
Source
Made by Hieu T. Luu. Original skill: brainstorm-coach, published on GitHub under the MIT licence.
View on GitHub