Analogical Reasoning MCP Server

A structured framework for constructing, mapping, and evaluating analogies to enhance systematic analogical thinking and problem-solving.

Core Concepts

Domain Elements

Domain elements are the building blocks of analogical reasoning. Each element has:

• A unique identifier
• A name and type (entity, attribute, relation, process)
• A descriptive explanation

Example:

{
  "id": "water_flow",
  "name": "Water Flow",
  "type": "process",
  "description": "Movement of water through pipes under pressure"
}

Analogical Mappings

Mappings define correspondences between source and target domain elements. They include:

• Source and target element references
• Mapping strength (0.0-1.0)
• Justification for the mapping
• Known limitations

Example:

{
  "sourceElement": "water_flow",
  "targetElement": "electric_current",
  "mappingStrength": 0.9,
  "justification": "Both involve flow of substance through conduits",
  "limitations": ["Water is visible, electricity is not"]
}

Inferences

Inferences are conclusions drawn from analogical mappings. They contain:

• Statement of the inference
• Confidence level (0.0-1.0)
• Supporting mappings that justify the inference

Example:

{
  "statement": "Electrical resistance is like pipe friction",
  "confidence": 0.8,
  "basedOnMappings": ["water_flow_to_current", "pipe_to_wire"]
}

API

Tools

• analogicalReasoning
  • Construct and evaluate analogical mappings between domains
  • Input: Comprehensive analogical reasoning data structure
    • sourceDomain (object): Source domain with name and elements
    • targetDomain (object): Target domain with name and elements
    • mappings (array): Analogical mappings between domains
    • analogyId (string): Unique identifier for the analogy
    • purpose (enum): "explanation" | "prediction" | "problem-solving" | "creative-generation"
    • confidence (number): Overall confidence in the analogy (0.0-1.0)
    • iteration (number): Current iteration of the analogical reasoning process
    • strengths (string[]): Areas where the analogy is particularly strong
    • limitations (string[]): Known limitations of the analogy
    • inferences (array): Conclusions drawn from the mappings
    • nextOperationNeeded (boolean): Whether further operations are required
    • suggestedOperations (array): Recommended next steps
  • Returns structured analogical analysis with mappings and evaluations
  • Supports iterative refinement of analogical reasoning

Setup

Cursor

Add to Cursor

Or manually add the following to your ~/.cursor/mcp.json file:

{
  "mcpServers": {
    "Analogical Reasoning": {
      "command": "bunx",
      "args": ["@wemake.cx/analogical-reasoning@latest"]
    }
  }
}

Raycast

Add to Raycast

Or manually via UI by using Manage MCP Servers, and press CMD + N and paste the following:

{
  "mcpServers": {
    "Analogical Reasoning": {
      "command": "bunx",
      "args": ["@wemake.cx/analogical-reasoning@latest"]
    }
  }
}

System Prompt

The prompt for utilizing analogical reasoning should encourage systematic mapping and evaluation:

Follow these steps for analogical reasoning:

1. Domain Definition:
   - Clearly define both source and target domains
   - Identify key entities, attributes, relations, and processes in each domain
   - Ensure domains are well-structured before proceeding

2. Systematic Mapping:
   - Create explicit mappings between corresponding elements
   - Assign mapping strength based on structural similarity
   - Provide clear justification for each mapping
   - Identify limitations where mappings break down

3. Inference Generation:
   - Draw conclusions based on established mappings
   - Assign confidence levels to inferences
   - Reference supporting mappings for each inference
   - Consider alternative interpretations

4. Evaluation and Refinement:
   - Assess overall analogy quality and limitations
   - Identify areas for improvement or alternative source domains
   - Iterate on mappings based on new insights
   - Document lessons learned for future analogical reasoning