Brainstorming Agent Skill
Collaborative design refinement through Socratic dialogue. Use when exploring ideas, designing features, or planning architecture. Not for mechanical implementation tasks.
Install
one command, takes just this skill from the repositorynpx skills add https://github.com/majiayu000/spellbook --skill brainstormingThe instruction itself
13 sections, as written by the authorBrainstorming
> From obra/superpowers
Purpose
Turn rough ideas into fully-formed designs through collaborative dialogue, NOT immediate implementation.
Use for: Exploring ideas, designing features, architecture decisions
NOT for: Clear mechanical processes with obvious implementation
The Three Phases
Phase 1: Understanding the Idea
- Review existing context
- Read relevant project files
- Understand current architecture
- Note constraints and requirements
- Ask clarifying questions
- ONE question at a time
- Prefer multiple choice when possible
- Open-ended questions are fine too
- Wait for answer before next question
Good: "Should this feature support offline mode?
A) Yes, full offline support
B) Yes, limited offline (read-only)
C) No, online only"
Bad: "What about offline support? And caching?
Also, how should errors be handled?"
Phase 2: Exploring Approaches
- Present 2-3 different solutions
- Each with clear trade-offs
- Lead with recommended option
- Explain reasoning
## Approach A: Event-Driven (Recommended)
- Pros: Decoupled, scalable, easy to extend
- Cons: More complex debugging, eventual consistency
- Best for: Systems needing flexibility
## Approach B: Direct Integration
- Pros: Simple, immediate consistency
- Cons: Tight coupling, harder to scale
- Best for: Small, stable systems
## Approach C: Hybrid
- Pros: Balance of both
- Cons: Complexity of managing two patterns
- Best for: Gradual migration scenarios
- Discuss trade-offs
- Performance vs simplicity
- Flexibility vs maintainability
- Short-term vs long-term
Phase 3: Presenting the Design
- Present in sections (200-300 words each)
- Architecture overview
- Key components
- Data flow
- Error handling
- Testing strategy
- Validate incrementally
- Get confirmation after each section
- Adjust based on feedback
- Don't dump everything at once
Key Principles
Ask Questions Sequentially
❌ Wrong:
"What's the expected load? How many users?
What about mobile? Should it work offline?
What's the budget? Timeline?"
✅ Right:
"What's the expected user load?
A) <100 concurrent users
B) 100-1000 concurrent users
C) 1000+ concurrent users"
[Wait for answer]
"Given that load, should we prioritize...?"
YAGNI Ruthlessly
Remove unnecessary features from ALL designs:
- Don't add "just in case" features
- Don't over-engineer for hypothetical scale
- Don't add flexibility that isn't needed yet
Explore Before Deciding
Never jump to first solution:
- Consider at least 2-3 approaches
- Understand trade-offs of each
- Make informed recommendation
Present Incrementally
Don't dump full design at once:
- Section by section
- Validate understanding
- Adjust as needed
Follow-up Actions
After design is validated:
- Document the design
- Create design doc or ADR
- Include key decisions and rationale
- Commit to version control
- Design docs belong in repo
- Track changes over time
- Move to implementation planning
- Break into tasks
- Estimate effort
- Identify dependencies
Example Session
User: "I want to add notifications to the app"
Other skills for the same job
different authors, same section of the catalogue
Guide users through a structured workflow for co-authoring documentation. Use when user wants to write documentation, proposals, technical specs, decision docs, or similar structured content. This workflow helps users efficiently transfer context, refine content through iteration, and verify the doc works for readers. Trigger when user mentions writing docs, creating proposals, drafting specs, or similar documentation tasks.
Intelligently organizes your files and folders across your computer by understanding context, finding duplicates, suggesting better structures, and automating cleanup tasks. Reduces cognitive load and keeps your digital workspace tidy without manual effort.
Generates creative domain name ideas for your project and checks availability across multiple TLDs (.com, .io, .dev, .ai, etc.). Saves hours of brainstorming and manual checking.
You MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent, requirements and design before implementation.
Implements Manus-style file-based planning for complex tasks. Creates task_plan.md, findings.md, and progress.md. Use when starting complex multi-step tasks, research projects, or any task requiring >5 tool calls.
Creative research ideation and exploration. Use for open-ended brainstorming sessions, exploring interdisciplinary connections, challenging assumptions, or identifying research gaps. Best for early-stage research planning when you do not have specific observations yet. For formulating testable hypotheses from data use hypothesis-generation.
Comprehensive GitHub project management with swarm-coordinated issue tracking, project board automation, and sprint planning
Interview the user relentlessly about a plan or design until reaching shared understanding, resolving each branch of the decision tree. Use when user wants to stress-test a plan, get grilled on their design, or mentions "grill me".
How to use it
Take majiayu000/brainstorming from the repository into ~/.claude/skills for personal
use, or into .claude/skills inside a project.
The agent identifies a skill by the name field in its header. Two skills with the
same name cannot sit side by side — one of them will be ignored.