Gemini MCP

Instructions

Implementation Reference

• src/tools/business-tools.js:20-134 (handler)

  The asynchronous handler function that executes the tool's logic: validates inputs, generates AI prompts for financial analysis and executive summary, calls AI client, stores results in storage, and returns formatted output with ROI analysis.

      handler: async (args) => {
        const { decision, context = '', timeline = '6 months', team_size = 5, risk_tolerance = 'medium' } = args;
        validateString(decision, 'decision');
        
        const timer = performanceMonitor.startTimer('financial_impact');
        
        const analysisPrompt = `Perform comprehensive financial impact analysis for this technical decision:
  
  **Decision**: ${decision}
  **Context**: ${context}
  **Timeline**: ${timeline}
  **Team Size**: ${team_size}
  **Risk Tolerance**: ${risk_tolerance}
  
  Provide detailed financial analysis:
  
  1. **Cost Analysis**
     - Development costs (labor, tools, infrastructure)
     - Implementation costs (deployment, training, migration)
     - Ongoing operational costs (maintenance, support, updates)
     - Hidden costs (integration, testing, documentation)
  
  2. **Benefit Analysis**
     - Performance improvements (quantified value)
     - Developer productivity gains
     - Maintenance cost reductions
     - Risk mitigation value
     - Competitive advantages
  
  3. **ROI Calculation**
     - Total Cost of Ownership (TCO) over 3 years
     - Expected return on investment
     - Payback period calculation
     - Net present value (NPV) analysis
  
  4. **Risk Assessment**
     - Financial risks and mitigation costs
     - Probability-weighted scenarios
     - Worst-case financial impact
     - Risk-adjusted ROI
  
  5. **Decision Framework**
     - Go/No-go recommendation with reasoning
     - Alternative options analysis
     - Phased implementation strategy
     - Success metrics and KPIs
  
  Provide specific dollar amounts where possible and explain assumptions.`;
  
        const financialAnalysis = await aiClient.call(analysisPrompt, 'analysis', { 
          complexity: 'complex',
          maxTokens: 4000 
        });
        
        // Generate executive summary
        const executivePrompt = `Create an executive summary for C-suite consumption:
  
  ${financialAnalysis}
  
  Format as executive briefing with:
  1. **Investment Summary** (one paragraph)
  2. **Key Financial Metrics** (ROI, Payback, NPV)
  3. **Strategic Recommendation** (Proceed/Pause/Pivot)
  4. **Risk Mitigation** (top 3 risks and solutions)
  5. **Success Measures** (how to track ROI)
  
  Use business language, avoid technical jargon.`;
  
        const executiveSummary = await aiClient.call(executivePrompt, 'analysis');
        
        // Save financial analysis for tracking
        const analysisData = {
          id: Date.now().toString(),
          decision,
          context,
          timeline,
          team_size,
          risk_tolerance,
          timestamp: new Date().toISOString(),
          analysis: financialAnalysis,
          executive_summary: executiveSummary
        };
        
        const storageData = await storage.read('financial_analyses');
        if (!storageData.analyses) storageData.analyses = [];
        storageData.analyses.push(analysisData);
        
        // Keep only last 100 analyses
        if (storageData.analyses.length > 100) {
          storageData.analyses = storageData.analyses.slice(-100);
        }
        
        await storage.write('financial_analyses', storageData);
        
        timer.end();
        
        return `💰 **Financial Impact Analysis**
  
  **Decision**: ${decision}
  **Timeline**: ${timeline} | **Team**: ${team_size} people | **Risk Tolerance**: ${risk_tolerance}
  
  ---
  
  📊 **Executive Summary**
  
  ${executiveSummary}
  
  ---
  
  📈 **Detailed Financial Analysis**
  
  ${financialAnalysis}
  
  **Analysis ID**: ${analysisData.id} (saved for portfolio tracking)`;
      }

• src/tools/business-tools.js:11-19 (schema)

  The tool definition object including name, description, and input parameters schema used for validation.

  'mcp__gemini__financial_impact': {
    description: 'ROI analysis and cost-benefit calculations for technical decisions with business impact quantification',
    parameters: {
      decision: { type: 'string', description: 'Technical decision to analyze', required: true },
      context: { type: 'string', description: 'Business context and constraints' },
      timeline: { type: 'string', description: 'Implementation timeline', default: '6 months' },
      team_size: { type: 'number', description: 'Team size for implementation', default: 5 },
      risk_tolerance: { type: 'string', description: 'Risk tolerance level', default: 'medium' }
    },

• src/tools/registry.js:235-240 (registration)

  Registration of the businessTools module (containing mcp__gemini__financial_impact) into the central ToolRegistry via registerToolsFromModule.

  this.registerToolsFromModule(codeTools);
  this.registerToolsFromModule(analysisTools);
  this.registerToolsFromModule(enhancedTools);
  this.registerToolsFromModule(businessTools);
  this.registerToolsFromModule(licenseTools);

• src/tools/registry.js:8-8 (registration)

  Import of business-tools.js module containing the tool definition.

  import { businessTools } from './business-tools.js';