Solution Brainstorming

Solution Brainstorming

Overview

Brainstorming is the bridge from problem space to solution space. But unstructured brainstorming produces平庸 ideas. This skill provides structured ideation techniques — divergent thinking to generate possibilities, convergent thinking to select the most promising ones — all within constraints that make the process productive.

When to Use

• You've defined a problem and need solution options
• The obvious solutions are uninspired or already tried
• You want to involve the team in creative problem-solving
• You need to explore multiple approaches before committing
• Don't use for: well-known solutions to well-known problems (just build it), individual work (brainstorming works best with 2-5 people)

Instructions

Phase 1: Set Constraints (10 min)

Before generating ideas, define what's in bounds:

• Time: "We have one sprint to ship something. What can we build?"
• Resources: "We have one designer and two engineers."
• Quality bar: "It needs to work reliably but doesn't need to be beautiful."
• Must-nots: What's off the table?

Phase 2: Divergent Thinking (20 min)

Use these techniques to generate ideas:

SCAMPER:

• Substitute: What can we replace?
• Combine: What can we merge?
• Adapt: What existing solution can we borrow?
• Modify: What can we change?
• Put to other use: Who else could use this?
• Eliminate: What can we remove?
• Reverse: What's the opposite?

Crazy 8s: Each person folds paper into 8 sections. Draw one idea per section, 60 seconds each. 8 ideas in 8 minutes.

Worst Idea First: What's the WORST way to solve this? Then invert it — the opposite is often creative.

Analogy Thinking: "How would [Amazon/Uber/Apple/Notion] solve this?"

Phase 3: Capture Everything

• One idea per sticky note / card
• No judgment during divergence
• Wild ideas encouraged (they often spark practical ones)
• Aim for 20-40 ideas total

Phase 4: Convergent Thinking (15 min)

1. Cluster related ideas into themes
2. Dot vote: Each person gets 3-5 votes
3. Top ideas get 3+ votes → move to prioritization
4. Capture runner-ups — they might be useful later

Phase 5: Next Steps

For the top ideas:

1. Briefly describe each at a high level
2. Identify what assumptions need testing
3. Assign a "champion" to explore each idea further
4. Schedule a follow-up to converge on final selection

Sample Output

Problem: Users don't reach the aha moment in our onboarding

Technique: SCAMPER

• Substitute: Replace tutorial with interactive product tour
• Combine: Onboarding + first workflow = "learn by doing"
• Adapt: Duolingo-style progressive disclosure
• Eliminate: Remove signup, let them use the product first
• Reverse: Don't onboard users — onboard their data first

Top 3 ideas: (1) Learn-by-doing template, (2) No-signup sandbox, (3) Personalized onboarding based on role

Common Pitfalls

1. Judging too early — The "yes, but..." reflex kills creativity. Keep divergence and convergence separate
2. Too many constraints — No constraints = no focus. Too many constraints = no creativity. Find the balance
3. Dominant voices — The loudest person's ideas shouldn't win. Use dot voting or anonymous submission
4. Skipping the "bad" ideas — The worst idea often contains the seed of the best idea
5. No follow-through — Brainstorming without prioritization and action is just an interesting conversation

Verification Checklist

• [ ] Constraints defined before ideation
• [ ] Divergent and convergent phases kept separate
• [ ] 20+ ideas generated
• [ ] Top ideas identified with dot voting
• [ ] Next steps assigned for each top idea

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: uthumany
• Source: uthumany/uthy-legacy-os
• License: MIT
• Homepage: https://uthumany.github.io/uthy-legacy-os/

Install and usage instructions live in the source repository linked above.