brainstorming-partner
Brainstorming Partner
Overview
Brainstorming Partner is an interactive skill that helps you generate, evaluate, and refine ideas using multiple proven brainstorming frameworks. Instead of brainstorming alone, you get a reliable AI partner that applies six distinct frameworks to your challenge, evaluates each idea's feasibility, impact, and novelty, and presents ranked results for iterative refinement.
Use this skill when you need to:
- Generate innovative ideas for products, features, or services
- Solve complex business or technical problems
- Explore multiple perspectives on a challenge
- Evaluate ideas before committing resources
- Have structured idea-generation conversations with AI assistance
Brainstorming Workflow
1. Present Your Challenge
Start by describing what you want to brainstorm. Provide:
- Topic or Challenge: The problem, opportunity, or area you want to explore
- Context: Any constraints, goals, or background relevant to the brainstorming
- Desired Outcomes: What success looks like (optional but helpful)
Example: "I need to brainstorm how to reduce our 40% onboarding drop-off rate. Our users get frustrated during setup. What could we explore?"
2. Multi-Framework Idea Generation
The skill applies six complementary frameworks to generate diverse ideas:
- SCAMPER - Systematic innovation through substitution, combination, adaptation, modification, repurposing, elimination, and reversal
- Six Thinking Hats - Multi-perspective thinking (facts, emotions, risks, benefits, creativity, process)
- Mind Mapping - Hierarchical exploration of problems, stakeholders, solutions, and connections
- Lateral Thinking - Breaking patterns through random word association and unexpected connections
- Divergence-Convergence - Pure generation followed by structured evaluation
- Forced Connections - Combining your challenge with concepts from unrelated domains
See references/frameworks.md for detailed framework descriptions.
3. Idea Evaluation
Each generated idea is evaluated on three dimensions:
- Feasibility (0-10): How practical and actionable is it? Can you actually do it?
- Impact (0-10): How significant and valuable would it be if implemented?
- Novelty (0-10): How original is it? What's the degree of newness?
Composite Score = (Feasibility + Impact + Novelty) / 3
See references/evaluation_metrics.md for detailed rubrics and scoring guidance.
4. Ranked Results & Interactive Iteration
Ideas are ranked by composite score and presented with:
- Individual metric scores with reasoning
- Overall assessment (Strong/Good/Fair/Weak)
- Framework source (which technique generated it)
- Full description
You then iterate: ask for clarification, request refinement, explore specific ideas deeper, or generate new ideas based on feedback.
Using the Skill with Agents
Running Idea Generation
Execute the idea generation script:
python scripts/generate_ideas.py "Your brainstorming topic" \
--context "Additional context or constraints" \
--output ideas.json
This generates idea prompts across all six frameworks organized by type.
Evaluating Ideas
Run the evaluation script to score ideas on Feasibility, Impact, and Novelty:
python scripts/evaluate_ideas.py ideas.json \
--context "Your context" \
--topic "Your topic" \
--output evaluated_ideas.json
This produces:
- Individual scores for each metric
- Composite rankings
- Score distribution statistics
- Assessment summary for each idea
Testing the Skill
Test prompts are available in evals/evals.json covering:
- Product feature brainstorming
- Business problem-solving
- Creative campaign ideation
- Technical architecture decisions
- Content strategy
- User experience improvement
Evaluation Metrics Reference
Feasibility Scoring
How practical is the idea given real-world constraints?
- 9-10: Can implement immediately with existing resources
- 7-8: Achievable with some additional resources or effort
- 5-6: Possible but requires significant effort or new skills
- 3-4: Very difficult; requires major changes or external dependencies
- 1-2: Extremely difficult or unrealistic
- 0: Impossible or violates fundamental constraints
Impact Scoring
How valuable and transformative would it be?
- 9-10: Transformational—solves major problems or creates substantial value
- 7-8: Significant—meaningfully improves the situation
- 5-6: Moderate—provides measurable benefits
- 3-4: Marginal—provides minor improvements
- 1-2: Minimal—barely noticeable impact
- 0: No positive impact or negative effects
Novelty Scoring
How original and new is the idea?
- 9-10: Highly unique—truly novel in this space
- 7-8: Original—significant variation on existing approaches
- 5-6: Somewhat new—adaptation with new elements
- 3-4: Incremental—minor variations on standard approaches
- 1-2: Minimal newness—obvious or widely-used
- 0: Not new—direct copy of existing solutions
Composite Score Interpretation
- 8-10: Strong idea—pursue aggressively
- 6-7.9: Good idea—consider with conditions
- 5-5.9: Fair idea—useful for specific contexts
- 3-4.9: Weak idea—needs refinement
- 0-2.9: Poor idea—unlikely to pursue
Brainstorming Frameworks
SCAMPER
Systematically applies seven techniques to existing solutions:
- Substitute: Replace components or attributes
- Combine: Merge with other products or ideas
- Adapt: Adjust for other purposes
- Modify: Change form, scale, or attributes
- Put to other uses: Repurpose or use differently
- Eliminate: Remove or simplify
- Reverse: Invert order or direction
Best for: Product improvements, process optimization, features addition
Six Thinking Hats
Uses six thinking modes as "hats" for parallel thinking:
- White Hat: Facts and data—what do we know?
- Red Hat: Emotions and intuition—what do we feel?
- Black Hat: Critical thinking—what could go wrong?
- Yellow Hat: Optimism—what's the best case?
- Green Hat: Creativity—what else could work?
- Blue Hat: Process and organization—how do we proceed?
Best for: Complex problems, team decision-making, fair evaluation
Mind Mapping
Creates hierarchical structure showing relationships:
- Central theme with primary and secondary branches
- Shows problem aspects, stakeholders, solutions, and connections
- Identifies gaps and relationships visually
Best for: Breaking down complex topics, exploring relationships, identifying gaps
Lateral Thinking
Uses random stimuli to break habitual patterns:
- Force connections between unrelated concepts and your challenge
- Explore unexpected associations
- Follow tangential paths to novel solutions
Best for: Breaking mental blocks, truly novel ideas, creative problem-solving
Divergence-Convergence
Separates generation from evaluation:
- Divergence: Generate many ideas without judgment
- Convergence: Evaluate and filter ideas systematically
Best for: Encouraging quantity and novelty, preventing premature judgment
Forced Connections
Combines unrelated concepts and domains:
- List key attributes of your challenge
- Combine with concepts from different fields
- Explore implications of combinations
Best for: Cross-disciplinary innovation, unique positioning, product mashups
Iterative Refinement
After evaluation, iterate with your partner:
- Dive Deeper: Ask for more details on high-scoring ideas
- Combine Ideas: Merge complementary ideas from different frameworks
- Refine Weak Ideas: Explore how to improve lower-scoring ideas
- Explore Trade-offs: Understand feasibility vs. impact trade-offs
- Generate More: Request additional ideas in specific directions
- Prioritize: Decide which ideas to pursue based on your strategy
Example conversation:
- "Can you elaborate on the top three ideas?"
- "How could we make the high-novelty but low-feasibility ideas more doable?"
- "Which combination of these ideas would be most impactful?"
- "Generate more ideas focused specifically on reducing complexity"
Core Capabilities
1. Multi-Framework Idea Generation
Applies six proven brainstorming frameworks to generate diverse perspectives on your challenge.
2. Feasibility-Impact-Novelty Evaluation
Scores each idea on three dimensions to provide balanced assessment of viability and value.
3. Ranked Presentation
Ideas ordered by composite score with detailed reasoning for each metric.
4. Framework Transparency
Each idea tagged with source framework so you understand which technique generated it.
5. Interactive Iteration
Refine ideas, combine approaches, and generate new ideas based on feedback.
6. Contextual Adaptation
Frameworks applied flexibly to any brainstorming scenario: products, processes, content, strategy, technical decisions, etc.
Scripts Reference
scripts/generate_ideas.py
Generates brainstorming ideas using all six frameworks.
Usage:
python generate_ideas.py "Topic or challenge" \
--context "Additional context" \
--output ideas.json
Output: JSON file with ideas organized by framework.
Parameters:
topic(required): Your brainstorming topic--context: Additional constraints or background--output: Output file path (default: ideas.json)
scripts/evaluate_ideas.py
Evaluates ideas on Feasibility, Impact, and Novelty metrics.
Usage:
python evaluate_ideas.py ideas.json \
--context "Your context" \
--output evaluated_ideas.json \
--min-score 5.0
Output: JSON file with scored and ranked ideas.
Parameters:
input_file(required): Ideas JSON from generate_ideas.py--context: Context for evaluation--topic: Original brainstorming topic--output: Output file path (default: evaluated_ideas.json)--min-score: Minimum composite score to include (default: 0)
How to Brainstorm Effectively
Preparation
- Be Specific: The clearer your challenge, the better the ideas
- Provide Context: Share constraints, goals, and background
- Set Expectations: What does success look like?
During Brainstorming
- Suspend Judgment: Don't dismiss ideas immediately
- Explore All Frameworks: Each generates different idea types
- Look for Patterns: Which frameworks generate your best ideas?
- Combine Ideas: Often the best solution merges multiple ideas
After Evaluation
- Consider Trade-offs: High feasibility might mean lower novelty (and vice versa)
- Think About Portfolio: Mix of safe bets + moonshots
- Look for Quick Wins: High feasibility + high impact = immediate action
- Identify Refinement: What would make lower-scoring ideas viable?
- Iterate: The second round often produces better results
Tips for Better Results
- More Context Helps: Include goals, constraints, past attempts, and success criteria
- Diverse Thinking: Frameworks generate different idea types; use multiple
- Combine Frameworks: Best ideas often come from merging insights from different frameworks
- Iterate Aggressively: Refine ideas based on feedback; rarely right the first time
- Document Reasoning: Understand why ideas scored high helps with implementation