tim-hub/brainstormer

Brainstormer

Generate diverse ideas by applying structured thinking frameworks to any problem. Each framework forces a different angle, reducing blind spots and expanding the solution space.

How to Use

When the user asks to brainstorm:

1. Clarify the problem - Restate the problem/goal in one sentence to confirm understanding.
2. Select frameworks - Pick 3-5 frameworks from the list below based on the problem type. Use more for open-ended problems, fewer for constrained ones.
3. Generate ideas per framework - Apply each selected framework, producing 2-4 ideas each.
4. Present results - Use the structured output format below.
5. Synthesize - End with a "Top Picks" section combining the strongest ideas across frameworks.

Thinking Frameworks

First Principles

Break the problem down to its fundamental truths. Remove all assumptions. Rebuild from scratch.

• Best for: Architecture decisions, challenging "we've always done it this way"

SCAMPER

Apply each lens to the existing solution or concept:

• Substitute - What component can be replaced?
• Combine - What can be merged together?
• Adapt - What can be borrowed from elsewhere?
• Modify - What can be enlarged, shrunk, or reshaped?
• Put to other use - Can it serve a different purpose?
• Eliminate - What can be removed entirely?
• Reverse - What if we flip the order or logic?
• Best for: Improving existing features, refactoring, UX iteration

Reverse Thinking (Inversion)

Ask "How would we make this fail?" or "What's the opposite of what we want?" Then invert those answers into solutions.

• Best for: Risk analysis, finding hidden failure modes, defensive design

Analogy Transfer

Find a solved problem in a different domain that shares structural similarities. Map the solution back.

• Best for: Novel problems, cross-domain innovation, finding proven patterns

Constraint Manipulation

Deliberately add or remove constraints to shift thinking:

• "What if we had unlimited time/budget?"
• "What if this had to ship tomorrow?"
• "What if we could only use 10% of the code?"
• Best for: Breaking deadlocks, scope negotiation, MVP definition

Five Whys

Ask "why" repeatedly to drill to root cause, then brainstorm solutions at the deepest level.

• Best for: Debugging, root cause analysis, problem reframing

Stakeholder Perspectives

Consider the problem through the eyes of different stakeholders:

• End user, Developer, Product manager, Business/revenue, Security/compliance, Operations/SRE
• Best for: Feature design, trade-off decisions, prioritization

Framework Selection Guide

| Problem Type | Recommended Frameworks | |---|---| | New feature / greenfield | First Principles, Analogy Transfer, Stakeholder Perspectives | | Improve existing solution | SCAMPER, Constraint Manipulation, Stakeholder Perspectives | | Debugging / incident | Five Whys, Reverse Thinking, Constraint Manipulation | | Architecture / design | First Principles, Reverse Thinking, Analogy Transfer | | Stuck / no ideas | Constraint Manipulation, SCAMPER, Reverse Thinking | | Strategy / prioritization | Stakeholder Perspectives, First Principles, Constraint Manipulation |

Output Format

Present ideas grouped by framework in this structure:

## Brainstorm: [Problem Statement]

### [Framework Name]
| # | Idea | Rationale | Effort | Impact |
|---|------|-----------|--------|--------|
| 1 | ... | ... | Low/Med/High | Low/Med/High |
| 2 | ... | ... | Low/Med/High | Low/Med/High |

### [Next Framework Name]
| # | Idea | Rationale | Effort | Impact |
|---|------|-----------|--------|--------|
| 1 | ... | ... | Low/Med/High | Low/Med/High |

---

### Top Picks
The strongest ideas across all frameworks, with brief justification:
1. **[Idea]** - Why this stands out
2. **[Idea]** - Why this stands out
3. **[Idea]** - Why this stands out

Guidelines

• Prefer quantity over perfection in the generation phase; filter in the synthesis.
• Flag any idea that carries significant risk or trade-off.
• If the problem is domain-specific, weave in relevant domain knowledge.
• Keep each idea description to 1-2 sentences. Elaborate only if the user asks.
• When ideas conflict, present both sides rather than picking one silently.