🧠 zrald3 - Sequential Thinking MCP Server

Advanced sequential thinking and reasoning MCP server that integrates data from both zrald1 (file analysis) and zrald2 (function logic) to perform comprehensive code intelligence analysis.

🎯 Overview

zrald3 is the third server in the zrald trilogy, designed to perform sophisticated reasoning and analysis by combining insights from:

• zrald1: File structure, relationships, and graph data
• zrald2: Function analysis, logic patterns, and complexity metrics

The server uses sequential thinking methodology to break down complex problems into manageable steps, providing evidence-based insights and actionable recommendations.

🔗 Dependencies

REQUIRED: Both zrald1 and zrald2 must be running before using zrald3.

# Start zrald1 first
npx -y zrald1

# Start zrald2 second  
npx -y zrald2

# Then start zrald3
npx -y zrald3

📦 Installation

Global Installation

npm install -g zrald3

Direct Usage

npx -y zrald3

🛠 MCP Tools

1. sequential_thinking

Perform comprehensive sequential thinking analysis using integrated data.

Parameters:

• user_query (required): The question or problem to analyze
• thinking_depth: 'shallow' | 'medium' | 'deep' (default: 'medium')
• focus_areas: Array of specific areas to focus on
• max_steps: Maximum thinking steps (default: 10)
• include_evidence: Include detailed evidence (default: true)

Example:

{
  "user_query": "Analyze the codebase for security vulnerabilities and performance issues",
  "thinking_depth": "deep",
  "focus_areas": ["security", "performance", "complexity"],
  "max_steps": 15,
  "include_evidence": true
}

2. validate_dependencies

Check if both zrald1 and zrald2 are available and accessible.

Example:

{}

3. get_thinking_summary

Get a formatted summary of the latest thinking session.

Parameters:

• format: 'brief' | 'detailed' | 'technical' (default: 'detailed')

4. analyze_with_context

Analyze a specific context using integrated data.

Parameters:

• context (required): The specific context to analyze
• intent: 'analyze' | 'debug' | 'optimize' | 'understand' | 'design' | 'solve'
• complexity: 'low' | 'medium' | 'high'

5. get_insights

Get insights from the latest thinking session with optional filtering.

Parameters:

• priority_filter: 'all' | 'high' | 'critical' (default: 'all')

📊 Resources

• thinking://latest - Latest thinking session data
• insights://latest - Latest insights from analysis
• recommendations://latest - Latest recommendations

🔄 Sequential Thinking Process

zrald3 follows a structured thinking methodology:

1. Initial Analysis - Gather and assess data from zrald1 and zrald2
2. Data Correlation - Find relationships between file structure and function logic
3. Pattern Analysis - Examine logic patterns and architectural insights
4. Insight Synthesis - Generate actionable insights from findings
5. Recommendation Generation - Formulate specific recommendations

Each step includes:

• Evidence collection from both servers
• Confidence scoring
• Reasoning documentation
• Progressive refinement

🎯 Use Cases

Code Quality Analysis

{
  "user_query": "Evaluate overall code quality and identify improvement areas",
  "thinking_depth": "deep",
  "focus_areas": ["complexity", "maintainability", "patterns"]
}

Security Assessment

{
  "user_query": "Analyze the codebase for potential security vulnerabilities",
  "thinking_depth": "deep",
  "focus_areas": ["security", "input-validation", "error-handling"]
}

Performance Optimization

{
  "user_query": "Find performance bottlenecks and optimization opportunities",
  "thinking_depth": "medium",
  "focus_areas": ["performance", "complexity", "algorithms"]
}

Architecture Review

{
  "user_query": "Review the current architecture and suggest improvements",
  "thinking_depth": "deep",
  "focus_areas": ["architecture", "relationships", "modularity"]
}

🔧 Configuration

MCP Server Configuration

{
  "mcpServers": {
    "zrald1": {
      "command": "npx",
      "args": ["-y", "zrald1"]
    },
    "zrald2": {
      "command": "npx", 
      "args": ["-y", "zrald2"]
    },
    "zrald3": {
      "command": "npx",
      "args": ["-y", "zrald3"]
    }
  }
}

📈 Output Format

Thinking Result

{
  "success": true,
  "thinking_result": {
    "session": {
      "id": "session-uuid",
      "steps": [...],
      "confidence": 0.85,
      "conclusion": "Analysis summary"
    },
    "insights": [...],
    "recommendations": [...],
    "nextSteps": [...]
  },
  "dependencies_used": {
    "zrald1_files": 5,
    "zrald2_functions": 12,
    "correlations": 3,
    "insights": 4
  }
}

Insights

{
  "id": "insight-uuid",
  "type": "risk" | "opportunity" | "pattern" | "anomaly",
  "description": "Insight description",
  "priority": "low" | "medium" | "high" | "critical",
  "confidence": 0.9,
  "evidence": [...],
  "actionable": true
}

Recommendations

{
  "id": "rec-uuid",
  "type": "improvement" | "fix" | "optimization" | "investigation",
  "description": "Recommendation description",
  "priority": "low" | "medium" | "high" | "critical",
  "effort": "low" | "medium" | "high",
  "impact": "low" | "medium" | "high",
  "rationale": "Why this recommendation is important"
}

🚀 Integration Workflow

1. Start Dependencies

   npx -y zrald1  # File analysis server
   npx -y zrald2  # Function logic server

2. Start zrald3

   npx -y zrald3  # Sequential thinking server

3. Validate Setup

   {
     "tool": "validate_dependencies"
   }

4. Perform Analysis

   {
     "tool": "sequential_thinking",
     "arguments": {
       "user_query": "Your analysis question"
     }
   }

🔍 Error Handling

If dependencies are missing:

{
  "success": false,
  "error": "zrald1 MCP server is required but not accessible. Please start zrald1 first: npx -y zrald1",
  "requirement": "zrald1_required",
  "dependencies_status": {
    "zrald1": false,
    "zrald2": true
  }
}

📝 License

MIT License - see LICENSE file for details.

🤝 Contributing

This is part of the zrald MCP server trilogy. Contributions welcome!

🔗 Related Projects

• zrald1: File analysis and graph processing MCP server
• zrald2: Function logic analysis MCP server
• zrald3: Sequential thinking and reasoning MCP server (this project)