Dineway Brainstorming

Brainstorming Ideas Into Designs

Overview

Help turn ideas into fully formed designs and specs through natural collaborative dialogue.

Start by understanding the current project context, then ask questions one at a time to refine the idea. Before asking each question, check whether you can safely answer it with the recommended assumption and continue without interrupting the user. Once you understand what you're building, present the design and get user approval, applying the same check before every approval request. After clearing the design gate, automatically invoke the appropriate follow-up skill when the original request includes downstream work.

Anti-Pattern: "This Is Too Simple To Need A Design"

Every project goes through this process. A todo list, a single-function utility, a config change — all of them. "Simple" projects are where unexamined assumptions cause the most wasted work. The design can be short (a few sentences for truly simple projects), but you MUST present it and clear its approval gate before implementation.

Checklist

You MUST create a task for each of these items and complete them in order:

1. Explore project context — check files, docs, recent commits
2. Ask clarifying questions — one at a time; before each question, check whether the recommended assumption lets you proceed without asking
3. Propose 2-3 approaches — with trade-offs and your recommendation
4. Present design — in sections scaled to their complexity; after each section, run the recommendation-first check before requesting approval
5. Continue downstream work — when the original request includes more work, invoke the appropriate follow-up skill automatically

Process Flow

digraph dineway_brainstorming {
    "Explore project context" [shape=box];
    "Question genuinely needs user?" [shape=diamond];
    "Ask clarifying questions" [shape=box];
    "Propose 2-3 approaches" [shape=box];
    "Present design sections" [shape=box];
    "Approval genuinely needs user?" [shape=diamond];
    "User approves design?" [shape=diamond];
    "Brainstorming complete" [shape=box];
    "Downstream work remains?" [shape=diamond];
    "Invoke follow-up skill and continue" [shape=doublecircle];
    "Return design and stop" [shape=doublecircle];

    "Explore project context" -> "Question genuinely needs user?";
    "Question genuinely needs user?" -> "Propose 2-3 approaches" [label="no, use recommendation"];
    "Question genuinely needs user?" -> "Ask clarifying questions" [label="yes"];
    "Ask clarifying questions" -> "Propose 2-3 approaches";
    "Propose 2-3 approaches" -> "Present design sections";
    "Present design sections" -> "Approval genuinely needs user?";
    "Approval genuinely needs user?" -> "Brainstorming complete" [label="no, use recommendation"];
    "Approval genuinely needs user?" -> "User approves design?" [label="yes"];
    "User approves design?" -> "Present design sections" [label="no, revise"];
    "User approves design?" -> "Brainstorming complete" [label="yes"];
    "Brainstorming complete" -> "Downstream work remains?";
    "Downstream work remains?" -> "Invoke follow-up skill and continue" [label="yes"];
    "Downstream work remains?" -> "Return design and stop" [label="no"];
}

A cleared design is the terminal state of brainstorming, not necessarily of the user's request. If downstream work remains, pass the design forward, invoke the most appropriate follow-up skill automatically, and continue. Stop only when the user requested design alone, no downstream work remains, or a decision genuinely requires the user. Do NOT automatically write a design document or invoke a planning skill unless the request requires one.

The Process

Understanding the idea:

• Check out the current project state first (files, docs, recent commits)
• Ask questions one at a time to refine the idea
• Before each question, check whether project evidence and a clearly recommended assumption are sufficient to proceed; ask only when the answer cannot be safely inferred and genuinely requires the user's judgment or authority
• Prefer multiple choice questions when possible, but open-ended is fine too
• Only one question per message - if a topic needs more exploration, break it into multiple questions
• Focus on understanding: purpose, constraints, success criteria

Exploring approaches:

• Propose 2-3 different approaches with trade-offs
• Present options conversationally with your recommendation and reasoning
• Lead with your recommended option and explain why
• Before asking the user to pick an option, check whether you can safely choose and carry forward the recommendation yourself

Presenting the design:

• Once you believe you understand what you're building, present the design
• Scale each section to its complexity: a few sentences if straightforward, up to 200-300 words if nuanced
• Ask after each section whether it looks right so far, but first check whether the recommended decision is safe to carry forward without user approval
• Cover: architecture, components, data flow, error handling, testing
• Be ready to go back and clarify if something doesn't make sense

Continuing after the design:

• Re-read the original request after clearing the design gate
• If implementation or another downstream task remains, invoke the most appropriate applicable follow-up skill and continue automatically
• Do not ask for confirmation solely because brainstorming has ended

Key Principles

• Recommendation-first checkpoint - Before every question or approval request, decide whether the recommended option is safe to choose for the user
• Ask only when necessary - Ask only when the checkpoint shows that the choice genuinely requires the user's information, judgment, risk tolerance, or authority
• One question at a time - Don't overwhelm with multiple questions
• Multiple choice preferred - Easier to answer than open-ended when user input is genuinely required
• YAGNI ruthlessly - Remove unnecessary features from all designs
• Explore alternatives - Always propose 2-3 approaches before settling
• Incremental validation - Present the design section by section and clear each approval gate through either the safe recommendation or explicit user approval
• Automatic handoff - Treat the design gate as a phase boundary and invoke the appropriate follow-up skill when downstream work remains
• Be flexible - Go back and clarify when something doesn't make sense