Skip to main content
Data Science & AImelodic-software

research

Research domain knowledge using MCP servers (perplexity, context7, firecrawl). Gathers best practices, regulatory requirements, and competitive insights.

Stars
74
Source
melodic-software/claude-code-plugins
Updated
2026-04-07
Slug
melodic-software--claude-code-plugins--research
View on GitHubRaw SKILL.md

// install — copy + paste into any project

mkdir -p .claude/skills && curl -fsSL https://raw.githubusercontent.com/melodic-software/claude-code-plugins/HEAD/plugins/requirements-elicitation/skills/research/SKILL.md -o .claude/skills/research.md

Drops the SKILL.md into .claude/skills/research.md. Works with Claude Code, Cursor, and any agent that loads SKILL.md files from .claude/skills/.

Research Command

Research domain knowledge using MCP servers to enrich requirements elicitation.

Usage

/requirements-elicitation:research "e-commerce checkout best practices"
/requirements-elicitation:research "GDPR compliance" --domain "user-data" --depth deep
/requirements-elicitation:research "authentication patterns" --focus regulatory
/requirements-elicitation:research "competitor features" --focus competitive

Arguments

Argument Required Description
topic Yes The topic to research
--domain No Domain name for organizing output
--depth No Research depth: shallow, moderate, deep (default: moderate)
--focus No Research focus: best-practices, regulatory, competitive, technical, all

Research Capabilities

Best Practices Research

  • Industry standards
  • Common patterns
  • Recommended approaches
  • Lessons learned

Regulatory Research

  • Compliance requirements
  • Legal obligations
  • Industry regulations
  • Audit requirements

Competitive Research

  • Competitor features
  • Market standards
  • Differentiation opportunities
  • Feature gaps

Technical Research

  • Library capabilities
  • Framework requirements
  • Integration patterns
  • Technology constraints

Workflow

Step 1: Parse Research Request

research_request:
  topic: "{from argument}"
  domain: "{from --domain}"
  depth: shallow|moderate|deep
  focus: "{from --focus or all}"

Step 2: Load Domain Research Skill

Invoke the requirements-elicitation:domain-research skill to load MCP patterns.

Step 3: Execute MCP Queries

Based on focus area:

Best Practices:

queries:
  - server: perplexity
    query: "{topic} best practices 2025"
  - server: perplexity
    query: "{topic} common patterns recommendations"

Regulatory:

queries:
  - server: perplexity
    query: "{regulation} requirements {industry}"
  - server: firecrawl
    action: scrape regulatory documentation

Competitive:

queries:
  - server: firecrawl
    action: search competitor features
  - server: perplexity
    query: "{industry} market leaders features"

Technical:

queries:
  - server: context7
    action: resolve library, get docs
  - server: perplexity
    query: "{technology} integration requirements"

Step 4: Synthesize Findings

Combine research results:

  • Extract key findings
  • Identify requirements implications
  • Note confidence levels
  • Flag items needing validation

Step 5: Save and Report

Save to .requirements/{domain}/research/ Display summary of findings.

Examples

Best Practices Research

/requirements-elicitation:research "e-commerce checkout optimization"

Output:

Researching: e-commerce checkout optimization
Depth: moderate
Focus: best-practices

Querying MCP servers...
  [perplexity] e-commerce checkout best practices 2025
  [perplexity] cart abandonment reduction techniques

Key Findings:

1. CHECKOUT FLOW
   - Guest checkout reduces abandonment by 30%
   - Progress indicators increase completion
   - Mobile-first design essential

2. PAYMENT
   - Multiple payment options required
   - Saved payment methods increase conversion
   - Clear security indicators build trust

3. PERFORMANCE
   - Checkout should complete in < 3 steps
   - Page load < 2 seconds critical
   - Real-time validation reduces errors

Derived Requirements (8):
  REQ-RES-001: System shall support guest checkout
  REQ-RES-002: System shall display checkout progress
  REQ-RES-003: System shall support multiple payment methods
  REQ-RES-004: System shall complete checkout in 3 steps or fewer
  ... (4 more)

Confidence: MEDIUM (needs stakeholder validation)

Saved to: .requirements/checkout/research/RES-20251225-170000.yaml

Regulatory Research

/requirements-elicitation:research "PCI-DSS" --focus regulatory --depth deep

Output:

Researching: PCI-DSS compliance
Depth: deep
Focus: regulatory

Querying MCP servers...
  [perplexity] PCI-DSS requirements payment processing
  [perplexity] PCI-DSS 4.0 changes requirements
  [firecrawl] scraping PCI Security Standards Council

Key Findings:

1. DATA PROTECTION (Requirement 3)
   - Never store CVV after authorization
   - Encrypt stored card data (AES-256)
   - Mask PAN when displayed

2. ACCESS CONTROL (Requirement 7-8)
   - Restrict access to need-to-know
   - Unique IDs for each user
   - MFA for administrative access

3. MONITORING (Requirement 10)
   - Log all access to cardholder data
   - Retain logs for 1 year minimum
   - Daily log review required

4. TESTING (Requirement 11)
   - Quarterly vulnerability scans
   - Annual penetration testing
   - Change detection mechanisms

Derived Requirements (15):
  REQ-RES-001: System shall not store CVV/CVC after authorization [MUST]
  REQ-RES-002: System shall encrypt stored card data using AES-256 [MUST]
  REQ-RES-003: System shall mask PAN displaying only last 4 digits [MUST]
  ... (12 more)

Confidence: HIGH (from official documentation)

Saved to: .requirements/payment/research/RES-20251225-171500.yaml

Competitive Research

/requirements-elicitation:research "inventory management software" --focus competitive

Output:

Researching: inventory management software competitors
Depth: moderate
Focus: competitive

Querying MCP servers...
  [firecrawl] searching inventory management software features
  [perplexity] inventory management market leaders 2025

Competitors Analyzed:

1. CompetitorA
   - Real-time inventory tracking
   - Multi-warehouse support
   - Barcode scanning
   - Automated reorder points

2. CompetitorB
   - AI demand forecasting
   - Supplier management
   - Integration marketplace
   - Mobile app

3. CompetitorC
   - Simple interface
   - Low-cost option
   - Basic reporting
   - Limited integrations

Feature Matrix:
  Feature               | A | B | C | Our Need?
  Real-time tracking    | ✓ | ✓ | ✓ | Table stakes
  Multi-warehouse       | ✓ | ✓ | - | Should have
  AI forecasting        | - | ✓ | - | Differentiator?
  Supplier management   | ✓ | ✓ | - | Should have
  Mobile app            | ✓ | ✓ | - | Should have

Derived Requirements (10):
  REQ-RES-001: System shall provide real-time inventory visibility [MUST]
  REQ-RES-002: System shall support multiple warehouse locations [SHOULD]
  REQ-RES-003: System shall integrate with barcode scanners [SHOULD]
  ... (7 more)

Confidence: LOW (based on public information)

Saved to: .requirements/inventory/research/RES-20251225-173000.yaml

Technical Research

/requirements-elicitation:research "React state management" --focus technical

Output:

Researching: React state management
Depth: moderate
Focus: technical

Querying MCP servers...
  [context7] resolving react library
  [context7] getting docs: state management
  [perplexity] React state management patterns 2025

Key Findings:

1. BUILT-IN OPTIONS
   - useState for local state
   - useReducer for complex state
   - Context API for global state (with caveats)

2. EXTERNAL LIBRARIES
   - Redux Toolkit (complex apps)
   - Zustand (simple, performant)
   - Jotai (atomic state)
   - React Query (server state)

3. RECOMMENDATIONS
   - Start simple (useState, Context)
   - Add complexity only when needed
   - Separate server state from client state

Technical Constraints:
  - React 18+ required for new patterns
  - SSR considerations for Next.js
  - Bundle size implications

Derived Requirements (5):
  REQ-RES-001: System shall use React 18 or later [CONSTRAINT]
  REQ-RES-002: System shall separate server state from client state [SHOULD]
  REQ-RES-003: System shall minimize bundle size impact of state library [SHOULD]
  ... (2 more)

Confidence: MEDIUM (verify with team)

Saved to: .requirements/frontend/research/RES-20251225-174500.yaml

Output Format

Saved YAML Structure

research_session:
  id: "RES-{timestamp}"
  topic: "{topic}"
  domain: "{domain}"
  depth: shallow|moderate|deep
  focus: "{focus area}"
  timestamp: "{ISO-8601}"

  queries:
    - server: perplexity
      query: "{query text}"
      success: true

    - server: context7
      library: "{library}"
      topic: "{topic}"
      success: true

  findings:
    category_1:
      - "{finding 1}"
      - "{finding 2}"

    category_2:
      - "{finding 3}"

  derived_requirements:
    - id: REQ-RES-001
      text: "{requirement}"
      source: research
      source_detail: "{specific source}"
      confidence: high|medium|low
      needs_validation: true
      priority: must|should|could

  recommendations:
    - "{recommendation 1}"
    - "{recommendation 2}"

  gaps_for_further_research:
    - "{topic needing more research}"

Integration

Follow-Up Commands

# Validate research with stakeholders
/requirements-elicitation:interview "Technical Lead" --context "validate React findings"

# Check if research fills gaps
/requirements-elicitation:gaps

# More research on specific topic
/requirements-elicitation:research "Zustand vs Redux" --focus technical --depth deep

# Consolidate all sources
/requirements-elicitation:discover "{domain}"