Agent skill

brainstorm

Collaborative brainstorming session on any topic — project-related or general. Use when asked to 'brainstorm', 'let's brainstorm', 'explore ideas', 'think through', or 'brainstorm about [topic]'. NOT for feature discovery (use /nf), NOT for PRD/JTBD docs (use /product), NOT for deep research (use /deep-research), NOT for pre-implementation design (auto-triggered by design-exploration skill).

Stars 163
Forks 31

Install this agent skill to your Project

npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/other/other/brainstorm-alexandrbasis-wythm-claude-workflo

SKILL.md

Brainstorming Session

Objective

Conduct a collaborative brainstorming session on any topic - project-related or general - through natural dialogue, exploration of options, and structured capture of insights.

Guidelines

  • Use AskUserQuestion tool for ALL clarifications - provides interactive options for user to choose from
  • Ask non-obvious and thought-provoking questions that challenge assumptions
  • Continue until the topic is fully explored - don't stop early
  • Present multiple perspectives and approaches
  • Capture key insights and decisions in the brainstorm notes

Workflow

Step 1: Determine Scope

Ask the user whether the topic is:

  • Project-related: Will explore codebase context first
  • General: Skip context gathering, proceed to exploration

Step 2: Context Phase (Project-Related Only)

If project-related, invoke the design-exploration skill which will:

  • Use multiple Explore agents in parallel (Sonnet) to gather project context
  • Identify relevant files, patterns, and existing implementations
  • Build understanding of constraints and opportunities

Step 3: Exploration Phase

For ALL topics (project-related or general):

Understanding the Topic:

  • Ask clarifying questions (batch related questions)
  • Prefer multiple choice when possible
  • Focus on: goals, constraints, success criteria, concerns

Exploring Approaches:

  • Propose 2-3 different perspectives or options
  • Present trade-offs clearly
  • Lead with your recommendation and reasoning

Deep Exploration - Ask non-obvious questions across categories:

Practical & Implementation:

  • What could go wrong? Edge cases?
  • What resources/dependencies are needed?
  • How would this scale or evolve?

Perspective & Assumptions:

  • What assumptions are we making?
  • Who else is affected by this?
  • What's the opposite approach look like?

Impact & Value:

  • How do we measure success?
  • What's the minimum viable version?
  • What happens if we don't do this?

Trade-offs:

  • Speed vs quality considerations
  • Short-term vs long-term implications
  • Complexity vs simplicity

Presenting Ideas:

  • Break into 200-300 word sections
  • Validate understanding after each section
  • Be ready to pivot if direction changes

Step 4: Research Phase (When Needed)

If exploration reveals topics requiring code context or external research:

Code Research (Project-Related):

  • Use Explore agents with Sonnet to scan relevant modules, patterns, and prior art
  • Capture constraints, existing APIs, and integration points

Quick Research (Exa MCP):

  • get_code_context_exa — for code-related context, APIs, libraries
  • web_search_exa — for trends, market data, best practices

Deep Research:

  • Launch comprehensive-researcher agent (Task tool) for topics requiring:
    • Multiple sources and cross-verification
    • Industry benchmarks or competitive analysis
    • Technical documentation deep-dives
  • Parallel research: Launch multiple researcher agents simultaneously for different topics

Launch research proactively when:

  • Topic requires current or up-to-date information
  • Market trends, competitor analysis, or industry standards are relevant
  • Technical decisions benefit from external validation
  • Exploring unfamiliar domains or technologies
  • Knowledge gaps are identified during exploration

No permission needed - launch research agents proactively when the conversation reveals a need. Inform user what's being researched and continue the brainstorming while agents work in background.

Step 5: Capture Phase

After exploration is complete:

  1. Create brainstorm notes: docs/brainstorming/brainstorm-YYYY-MM-DD-[topic-slug].md
  2. Use template: docs/product-docs/templates/brainstorm-template.md
  3. Include:
    • Topic overview and type (project/general)
    • Key questions explored
    • Options discussed with pros/cons
    • Conclusions and insights
    • Action items (if any)
  4. Present summary to user for confirmation

Step 6: Codex Validation (Project-Related Only, Optional)

For project-related brainstorms with significant scope, offer cross-AI validation.

AskUserQuestion: "Run cross-AI validation on brainstorm notes via /codex-cli?"

  • "Yes" — invoke /codex-cli for review
  • "Skip" — finalize without validation

If approved:

  • Invoke /codex-cli for brainstorm notes review as senior product analyst
  • Focus: completeness, feasibility, no conflicts with existing architecture
  • Update notes if needed, add "Cross-AI Validation: PASSED" when approved

Skip this step entirely for general (non-project) brainstorms.

Scope Boundaries

  • Feature discovery interviews: use /nf
  • PRD/JTBD documentation: use /product
  • Deep technical research: use /deep-research
  • Pre-implementation design: auto-triggered by design-exploration skill (not this one)
  • Task creation: use /ct

Output

docs/brainstorming/brainstorm-YYYY-MM-DD-[topic-slug].md

Expand your agent's capabilities with these related and highly-rated skills.

Didn't find tool you were looking for?

Be as detailed as possible for better results