Think Strategies

#!/usr/bin/env node /** * Session Enhancement Tool * * Analyzes existing thinking sessions and intelligently populates metadata: * - sessionPurpose: Extracted from session content and context * - qualityRating: Based on analysis of thinking effectiveness * - automaticMetrics: Calculated from session data * * Usage: * node enhance-session.js <session-id> --purpose "Custom purpose" --quality '{"usefulness": 4, ...}' * node enhance-session.js <session-id> --auto-analyze * node enhance-session.js <session-id> --interactive * * Examples: * node enhance-session.js linear-session-20250726-142148 --auto-analyze * node enhance-session.js react-session-20250726-001238 --purpose "MCP server workflow demonstration" */ import fs from 'fs-extra'; import path from 'path'; import os from 'os'; import { fileURLToPath } from 'url'; const __filename = fileURLToPath(import.meta.url); const __dirname = path.dirname(__filename); class SessionEnhancer { constructor(storagePath) { this.storagePath = storagePath || path.join(os.homedir(), 'Documents', 'thinking'); } async enhanceSession(sessionId, options = {}) { const sessionPath = path.join(this.storagePath, sessionId, 'session.json'); if (!await fs.pathExists(sessionPath)) { throw new Error(`Session not found: ${sessionId}`); } const sessionData = await fs.readJson(sessionPath); console.log(`\n🔍 Analyzing session: ${sessionId}`); console.log(`Strategy: ${sessionData.strategy}`); console.log(`Thoughts: ${sessionData.thoughtHistory.length}`); console.log(`Branches: ${Object.keys(sessionData.branches).length}`); // Create backup const backupPath = sessionPath.replace('.json', '.enhanced.backup'); await fs.copy(sessionPath, backupPath); let enhancements = {}; // Handle different enhancement modes if (options.autoAnalyze) { enhancements = await this.autoAnalyzeSession(sessionData); } else if (options.interactive) { enhancements = await this.interactiveEnhancement(sessionData); } else { // Manual enhancement from provided options if (options.purpose) enhancements.sessionPurpose = options.purpose; if (options.quality) enhancements.qualityRating = options.quality; } // Always recalculate automatic metrics enhancements.automaticMetrics = this.calculateAutomaticMetrics(sessionData); // Apply enhancements const enhancedData = { ...sessionData, ...enhancements }; await fs.writeJson(sessionPath, enhancedData, { spaces: 2 }); console.log(`\n✅ Enhanced session: ${sessionId}`); console.log(`📄 Backup created: ${path.basename(backupPath)}`); this.printEnhancements(enhancements); return enhancements; } async autoAnalyzeSession(sessionData) { console.log(`\n🤖 Auto-analyzing session content...`); const analysis = { sessionPurpose: this.extractPurpose(sessionData), qualityRating: this.calculateQualityRating(sessionData) }; return analysis; } extractPurpose(sessionData) { const thoughtHistory = sessionData.thoughtHistory; const strategy = sessionData.strategy; if (thoughtHistory.length === 0) { return `${strategy} strategy session (no content)`; } const firstThought = thoughtHistory[0].thought; const lastThought = thoughtHistory[thoughtHistory.length - 1].thought; // Extract key topics and intent const allText = thoughtHistory.map(t => t.thought).join(' ').toLowerCase(); // Look for explicit purpose statements const purposeIndicators = [ /(?:i want to|i need to|goal is to|purpose is to|analyzing|exploring|testing|implementing|designing|reviewing|debugging|planning|solving)[\s\w]+/gi, /(?:this session|this thinking|this analysis)[\s\w]+/gi ]; for (const indicator of purposeIndicators) { const matches = allText.match(indicator); if (matches && matches[0]) { let purpose = matches[0].charAt(0).toUpperCase() + matches[0].slice(1); if (purpose.length > 80) purpose = purpose.substring(0, 77) + '...'; return purpose; } } // Fallback: analyze content topics const topics = this.extractTopics(thoughtHistory); const primaryTopic = topics[0]?.term || 'unknown'; // Strategy-specific purpose generation const strategyPurposes = { 'linear': `Linear analysis of ${primaryTopic}`, 'chain_of_thought': `Step-by-step reasoning about ${primaryTopic}`, 'react': `Interactive exploration of ${primaryTopic}`, 'rewoo': `Planned execution workflow for ${primaryTopic}`, 'tree_of_thoughts': `Multi-path exploration of ${primaryTopic}`, 'step_back': `High-level analysis of ${primaryTopic}`, 'self_ask': `Question-driven investigation of ${primaryTopic}`, 'scratchpad': `Iterative calculation for ${primaryTopic}`, 'trilemma': `Balanced optimization of ${primaryTopic}`, 'self_consistency': `Consensus building on ${primaryTopic}` }; return strategyPurposes[strategy] || `${strategy} strategy session on ${primaryTopic}`; } calculateQualityRating(sessionData) { const thoughtHistory = sessionData.thoughtHistory; const strategy = sessionData.strategy; if (thoughtHistory.length === 0) { return null; // Can't rate empty sessions } // Base metrics calculation const thoughtCount = thoughtHistory.length; const hasConclusion = thoughtHistory.some(t => t.finalAnswer || !t.nextThoughtNeeded); const hasActions = thoughtHistory.some(t => t.plannedActions || t.actionResults); const hasRevisions = thoughtHistory.some(t => t.isRevision); const avgThoughtLength = thoughtHistory.reduce((sum, t) => sum + t.thought.length, 0) / thoughtCount; // Quality scoring (1-5 scale) let usefulness = 3; // Base score if (hasConclusion) usefulness += 1; if (thoughtCount >= 5) usefulness += 0.5; if (avgThoughtLength > 200) usefulness += 0.5; usefulness = Math.min(5, Math.max(1, Math.round(usefulness))); let effectiveness = 3; // Base score if (hasConclusion) effectiveness += 1; if (thoughtCount >= 3 && thoughtCount <= 10) effectiveness += 1; // Sweet spot if (thoughtCount > 15) effectiveness -= 1; // Potentially verbose effectiveness = Math.min(5, Math.max(1, Math.round(effectiveness))); let clarity = 3; // Base score if (avgThoughtLength > 100 && avgThoughtLength < 500) clarity += 1; // Good detail level if (hasConclusion) clarity += 1; clarity = Math.min(5, Math.max(1, Math.round(clarity))); let insights = 3; // Base score if (hasConclusion) insights += 1; if (thoughtCount >= 5) insights += 0.5; if (hasRevisions) insights += 0.5; // Shows reflection insights = Math.min(5, Math.max(1, Math.round(insights))); // Strategy fit based on usage patterns let strategyFit = this.calculateStrategyFit(strategy, thoughtHistory); let efficiency = 3; // Base score if (thoughtCount <= 5 && hasConclusion) efficiency += 2; // Concise and complete else if (thoughtCount <= 10 && hasConclusion) efficiency += 1; else if (thoughtCount > 20) efficiency -= 1; // Potentially inefficient efficiency = Math.min(5, Math.max(1, Math.round(efficiency))); let actionability = 3; // Base score if (hasActions) actionability += 1; if (hasConclusion) actionability += 1; actionability = Math.min(5, Math.max(1, Math.round(actionability))); const reflection = this.generateReflection(thoughtHistory, { usefulness, effectiveness, clarity, insights, strategyFit, efficiency, actionability }); return { usefulness, effectiveness, clarity, insights, strategyFit, efficiency, actionability, reflection }; } calculateStrategyFit(strategy, thoughtHistory) { const hasActions = thoughtHistory.some(t => t.plannedActions || t.actionResults); const hasObservations = thoughtHistory.some(t => t.observation); const hasBranching = thoughtHistory.some(t => t.branchId); const hasRevisions = thoughtHistory.some(t => t.isRevision); switch (strategy) { case 'react': return hasActions && hasObservations ? 5 : 3; case 'rewoo': return hasActions ? 5 : 3; case 'tree_of_thoughts': return hasBranching ? 5 : 3; case 'linear': return hasRevisions ? 5 : 4; // Linear allows flexibility case 'chain_of_thought': return thoughtHistory.length >= 3 ? 4 : 3; default: return 4; // Default good fit } } generateReflection(thoughtHistory, ratings) { const avgRating = Object.values(ratings).reduce((sum, val) => sum + val, 0) / Object.keys(ratings).length; const thoughtCount = thoughtHistory.length; const hasConclusion = thoughtHistory.some(t => t.finalAnswer || !t.nextThoughtNeeded); if (avgRating >= 4.5) { return "Excellent thinking session with clear reasoning and actionable outcomes."; } else if (avgRating >= 4) { return "Strong thinking session with good progression and useful insights."; } else if (avgRating >= 3.5) { return "Solid thinking session with room for more depth or clearer conclusions."; } else if (avgRating >= 3) { return "Basic thinking session that covered the topic but could be more thorough."; } else { return "Limited thinking session that would benefit from more development."; } } calculateAutomaticMetrics(sessionData) { const thoughtHistory = sessionData.thoughtHistory; if (!sessionData.timestamp) { return null; } // Calculate approximate duration (if not already set) const duration = sessionData.automaticMetrics?.duration || Math.max(1, Math.round(thoughtHistory.length * 2)); // Estimate 2 min per thought // Calculate iteration ratio const revisionCount = thoughtHistory.filter(t => t.isRevision).length; const iterationRatio = thoughtHistory.length > 0 ? Math.round((revisionCount / thoughtHistory.length) * 100) / 100 : 0; // Check tool integration const toolIntegration = thoughtHistory.some(t => t.plannedActions || t.actionResults); return { duration, iterationRatio, toolIntegration }; } extractTopics(thoughtHistory) { const allText = thoughtHistory.map(t => t.thought).join(' '); const stopWords = new Set(['the', 'a', 'an', 'and', 'or', 'but', 'in', 'on', 'at', 'to', 'for', 'of', 'with', 'by', 'from', 'up', 'about', 'into', 'through', 'during', 'before', 'after', 'above', 'below', 'between', 'among', 'this', 'that', 'these', 'those', 'i', 'you', 'he', 'she', 'it', 'we', 'they', 'me', 'him', 'her', 'us', 'them', 'my', 'your', 'his', 'her', 'its', 'our', 'their', 'am', 'is', 'are', 'was', 'were', 'be', 'been', 'being', 'have', 'has', 'had', 'do', 'does', 'did', 'will', 'would', 'could', 'should', 'may', 'might', 'must', 'can', 'shall']); const words = allText.toLowerCase() .replace(/[^\w\s]/g, ' ') .split(/\s+/) .filter(word => word.length > 2 && !stopWords.has(word)); const wordCounts = {}; words.forEach(word => { wordCounts[word] = (wordCounts[word] || 0) + 1; }); return Object.entries(wordCounts) .sort(([,a], [,b]) => b - a) .slice(0, 5) .map(([word, count]) => ({ term: word, frequency: count })); } printEnhancements(enhancements) { console.log(`\n📊 Applied Enhancements:`); if (enhancements.sessionPurpose) { console.log(`🎯 Purpose: ${enhancements.sessionPurpose}`); } if (enhancements.qualityRating) { const rating = enhancements.qualityRating; const avgScore = Object.values(rating).filter(v => typeof v === 'number') .reduce((sum, val, _, arr) => sum + val / arr.length, 0).toFixed(1); console.log(`⭐ Quality: ${avgScore}/5.0 average`); console.log(` - Usefulness: ${rating.usefulness}/5`); console.log(` - Effectiveness: ${rating.effectiveness}/5`); console.log(` - Clarity: ${rating.clarity}/5`); console.log(` - Insights: ${rating.insights}/5`); console.log(` - Strategy Fit: ${rating.strategyFit}/5`); console.log(` - Efficiency: ${rating.efficiency}/5`); console.log(` - Actionability: ${rating.actionability}/5`); console.log(`💭 Reflection: ${rating.reflection}`); } if (enhancements.automaticMetrics) { const metrics = enhancements.automaticMetrics; console.log(`🔢 Metrics:`); console.log(` - Duration: ${metrics.duration} minutes`); console.log(` - Iteration Ratio: ${(metrics.iterationRatio * 100).toFixed(0)}%`); console.log(` - Tool Integration: ${metrics.toolIntegration ? 'Yes' : 'No'}`); } } } // CLI handling async function main() { const args = process.argv.slice(2); if (args.length === 0) { console.log(` Usage: node enhance-session.js <session-id> [options] Options: --auto-analyze Automatically analyze and populate all metadata --interactive Interactive enhancement mode --purpose "text" Set specific purpose --quality '{"usefulness": 4, ...}' Set specific quality rating (JSON) --storage-path PATH Custom storage path (default: ~/Documents/thinking) Examples: node enhance-session.js linear-session-20250726-142148 --auto-analyze node enhance-session.js react-session-20250726-001238 --purpose "MCP workflow demo" `); process.exit(1); } const sessionId = args[0]; let options = {}; for (let i = 1; i < args.length; i++) { switch (args[i]) { case '--auto-analyze': options.autoAnalyze = true; break; case '--interactive': options.interactive = true; break; case '--purpose': options.purpose = args[++i]; break; case '--quality': try { options.quality = JSON.parse(args[++i]); } catch (e) { console.error('❌ Invalid JSON for quality rating'); process.exit(1); } break; case '--storage-path': options.storagePath = args[++i]; break; } } try { const enhancer = new SessionEnhancer(options.storagePath); await enhancer.enhanceSession(sessionId, options); console.log('\n🎉 Enhancement completed successfully!'); } catch (error) { console.error(`❌ Enhancement failed: ${error.message}`); process.exit(1); } } // Only run if this script is executed directly if (import.meta.url === `file://${process.argv[1]}`) { main(); } export default SessionEnhancer;