Research Brief

Generate a comprehensive research design with model-optimized prompts for multi-LLM research.

When to Use

Use this skill when you need to:

•Design a research strategy that leverages multiple AI models
•Decompose complex research questions into MECE sub-questions
•Generate prompts optimized for Claude, Gemini, and GPT strengths
•Plan systematic research with risk assessment

Workflow

Invoke the create-research-brief skill Phase 1 for: "$ARGUMENTS"

Default Parameters

Parameter	Value	Rationale
`research_type`	market	Default; override based on question
`risk_depth`	standard	Balanced risk assessment
`model_mode`	parallel	Run models simultaneously
`multi_hypothesis`	false	Enable for hypothesis-driven research
`expert_panel`	false	Enable for strategic decisions

Research Types

Select based on the research question:

•market - Market sizing, trends, demand analysis
•competitive - Competitor analysis, positioning, strategy
•technology - Tech evaluation, build vs buy, capabilities
•strategic - Strategic planning, scenario analysis, options

Model Assignment Strategy

Model	Strength	Best For
Claude Opus 4.5	Judgment, synthesis, nuance	Strategic questions, conflict resolution
Gemini Pro 3	Breadth, citations, grounding	Factual lookup, comprehensive sourcing
GPT-5.2 Deep	Recency, depth, exhaustiveness	Technical details, edge cases

MECE Decomposition

The skill will decompose your research question into 5 categories following the appropriate MECE pattern:

Market Research:

•Market Size & Growth
•Market Structure & Segmentation
•Demand Drivers & Customer Needs
•Supply & Competition
•Evolution & Trends

Competitive Intelligence:

•Product & Capabilities
•Customers & Positioning
•Go-to-Market Strategy
•Organization & Resources
•Strategy & Trajectory

Technology Evaluation:

•Capability & Performance
•Maturity & Ecosystem
•Fit & Integration
•Cost & Economics
•Risk & Governance

Strategic Research:

•Current State Assessment
•External Environment
•Strategic Options
•Stakeholder Considerations
•Implementation Requirements

Output Format

The research brief will include:

xml

<research-brief>
  <header>
    <id>[unique identifier]</id>
    <type>[market|competitive|technology|strategic]</type>
    <objective>[research question]</objective>
  </header>

  <mece-decomposition>
    <category name="[Category 1]">
      <question>[Sub-question 1.1]</question>
      <question>[Sub-question 1.2]</question>
    </category>
    <!-- ... more categories ... -->
  </mece-decomposition>

  <model-prompts>
    <prompt model="claude" category="[assigned categories]">
      [Optimized prompt for Claude]
    </prompt>
    <prompt model="gemini" category="[assigned categories]">
      [Optimized prompt for Gemini]
    </prompt>
    <prompt model="gpt" category="[assigned categories]">
      [Optimized prompt for GPT]
    </prompt>
  </model-prompts>

  <risk-assessment depth="[quick|standard|comprehensive]">
    [Risk factors and mitigations]
  </risk-assessment>

  <next-steps>
    1. Execute prompts in each model
    2. Collect responses
    3. Run /consolidate-research with outputs
  </next-steps>
</research-brief>

Quality Gates

• Research objective is clear and answerable
• MECE decomposition has no overlaps or gaps
• Each sub-question is independently researchable
• Model assignments match model strengths
• Prompts are optimized for each model's style
• Risk assessment appropriate to decision stakes