decision-framework

Decision Framework MCP Server

A detailed tool for structured decision analysis and rational choice. This tool helps models systematically evaluate options, criteria, and outcomes using multiple decision frameworks.

Core Concepts

Decision Components

The decision framework operates on several key components:

• Options: Available alternatives or choices
• Criteria: Standards for evaluating options
• Outcomes: Possible results with probabilities and values
• Stakeholders: People or groups affected by the decision
• Constraints: Limitations on the decision

Example option:

{
  "id": "option_a",
  "name": "Cloud Migration",
  "description": "Migrate existing infrastructure to cloud platform"
}

Decision Analysis Types

The server supports multiple analysis frameworks:

• Expected Utility: Probability-weighted value calculations
• Multi-Criteria: Weighted scoring across multiple criteria
• Maximin: Choose option with best worst-case outcome
• Minimax Regret: Minimize maximum regret across scenarios
• Satisficing: Find first option meeting minimum criteria

Evaluation Process

Decisions progress through structured stages:

1. Problem Definition: Clear statement of decision to be made
2. Options: Identification of available alternatives
3. Criteria: Definition of evaluation standards
4. Evaluation: Scoring options against criteria
5. Analysis: Application of decision framework
6. Recommendation: Final recommendation with justification

API

Tools

• decisionFramework
  • Systematic decision analysis and rational choice evaluation
  • Input: Comprehensive decision analysis data structure
    • decisionStatement (string): Clear statement of the decision to be made
    • options (array): Available options or alternatives
      • Each option contains:
        • id (string): Unique identifier
        • name (string): Option name
        • description (string): Detailed description
    • criteria (array): Criteria for evaluating options
      • Each criterion contains:
        • id (string): Unique identifier
        • name (string): Criterion name
        • description (string): Detailed description
        • weight (number): Importance weight (0.0-1.0)
        • evaluationMethod (enum): "quantitative" | "qualitative" | "boolean"
    • criteriaEvaluations (array): Evaluations of options against criteria
      • Each evaluation contains:
        • criterionId (string): Reference to criterion
        • optionId (string): Reference to option
        • score (number): Score (0.0-1.0)
        • justification (string): Reasoning for the score
    • possibleOutcomes (array): Possible outcomes with probabilities
      • Each outcome contains:
        • id (string): Unique identifier
        • description (string): Outcome description
        • probability (number): Likelihood (0.0-1.0)
        • optionId (string): Associated option
        • value (number): Utility value
        • confidenceInEstimate (number): Confidence level (0.0-1.0)
    • informationGaps (array): Missing information affecting the decision
    • stakeholders (string[]): Affected parties
    • constraints (string[]): Decision limitations
    • timeHorizon (string): Decision timeframe
    • riskTolerance (enum): "risk-averse" | "risk-neutral" | "risk-seeking"
    • analysisType (enum): Type of analysis to perform
    • stage (enum): Current stage of decision process
    • iteration (number): Current iteration number
    • nextStageNeeded (boolean): Whether further analysis is required
  • Returns structured decision analysis with recommendations
  • Supports iterative refinement through multiple stages

Setup

bunx

{
  "mcpServers": {
    "Decision Framework": {
      "command": "bunx",
      "args": ["@wemake.cx/decision-framework@latest"]
    }
  }
}

bunx with custom settings

The server can be configured using the following environment variables:

{
  "mcpServers": {
    "Decision Framework": {
      "command": "bunx",
      "args": ["@wemake.cx/decision-framework@latest"],
      "env": {
        "DECISION_MAX_OPTIONS": "10",
        "DECISION_MAX_CRITERIA": "15",
        "DECISION_MIN_CONFIDENCE": "0.1"
      }
    }
  }
}

• DECISION_MAX_OPTIONS: Maximum number of options per decision (default: 20)
• DECISION_MAX_CRITERIA: Maximum number of criteria per decision (default: 20)
• DECISION_MIN_CONFIDENCE: Minimum confidence threshold for outcomes (default: 0.0)

System Prompt

The prompt for utilizing decision framework should encourage systematic analysis:

Follow these steps for structured decision making:

1. Problem Definition:
   - Clearly articulate the decision to be made
   - Identify the decision maker and key stakeholders
   - Define the time horizon and constraints
   - Establish the context and background

2. Option Generation:
   - Brainstorm comprehensive list of alternatives
   - Include creative and unconventional options
   - Ensure options are mutually exclusive and collectively exhaustive
   - Document assumptions and feasibility constraints

3. Criteria Development:
   - Identify all relevant evaluation criteria
   - Assign appropriate weights based on importance
   - Choose suitable evaluation methods for each criterion
   - Validate criteria with stakeholders

4. Systematic Evaluation:
   - Score each option against every criterion
   - Provide clear justification for scores
   - Consider uncertainty and confidence levels
   - Document key assumptions and trade-offs

5. Analysis and Recommendation:
   - Apply appropriate decision analysis framework
   - Conduct sensitivity analysis on key parameters
   - Identify information gaps and their impact
   - Provide clear recommendation with rationale

6. Implementation Planning:
   - Consider implementation challenges and risks
   - Develop contingency plans for key uncertainties
   - Establish monitoring and review mechanisms
   - Plan for decision communication and buy-in