{ "name": "What-If Simulation Agent", "nodes": [ { "parameters": { "httpMethod": "POST", "path": "mycroft-whatif", "responseMode": "responseNode", "options": { "rawBody": false } }, "id": "64d8d5b2-1138-4d63-acbe-5ec918f70dc4", "name": "🎯 Webhook β€” Receive Simulation Request", "type": "n8n-nodes-base.webhook", "typeVersion": 2, "position": [ -2352, 96 ] }, { "parameters": { "jsCode": "// ─── MYCROFT: Input Parser & Validator ───────────────────────────────────────\nconst body = $input.first().json.body || $input.first().json;\n\n// Apply defaults and parse types\nconst input = {\n company: String(body.company || 'Unknown Company'),\n ticker: String(body.ticker || 'N/A').toUpperCase(),\n revenue_growth: parseFloat(body.revenue_growth ?? 0.15),\n sentiment_score: parseFloat(body.sentiment_score ?? 0.50),\n risk_appetite: String(body.risk_appetite || 'medium').toLowerCase(),\n time_horizon: String(body.time_horizon || '1year').toLowerCase(),\n current_price: parseFloat(body.current_price ?? 100),\n pe_ratio: parseFloat(body.pe_ratio ?? 25),\n debt_equity: parseFloat(body.debt_equity ?? 0.50),\n market_cap_billions: parseFloat(body.market_cap_billions ?? 10)\n};\n\n// Validation rules\nconst errors = [];\nif (input.revenue_growth < -1 || input.revenue_growth > 20) errors.push('revenue_growth must be between -1.0 and 20.0');\nif (input.sentiment_score < 0 || input.sentiment_score > 1) errors.push('sentiment_score must be 0–1');\nif (!['low','medium','high'].includes(input.risk_appetite)) errors.push('risk_appetite must be low | medium | high');\nif (input.pe_ratio < 0) errors.push('pe_ratio must be positive');\nif (input.debt_equity < 0) errors.push('debt_equity must be positive');\n\n// Normalise time horizon\nif (!['3month','1year','5year'].includes(input.time_horizon)) input.time_horizon = '1year';\n\nif (errors.length > 0) throw new Error('Validation Failed: ' + errors.join(' | '));\n\n// Generate unique simulation ID\nconst rand = Math.random().toString(36).substr(2, 6).toUpperCase();\nconst simulation_id = `SIM-${Date.now()}-${rand}`;\n\nreturn [{ json: { ...input, simulation_id, timestamp: new Date().toISOString() } }];" }, "id": "138c670c-857d-4d56-9315-1c1f55f8ad2c", "name": "πŸ” Parse & Validate Input", "type": "n8n-nodes-base.code", "typeVersion": 2, "position": [ -2112, 96 ] }, { "parameters": { "jsCode": "// ─── SIMULATION 1: Growth vs Sentiment Tradeoff ──────────────────────────────\n// Core Principle: If fundamentals and market sentiment diverge β†’ detect the conflict\nconst data = $input.first().json;\nconst growth = data.revenue_growth;\nconst sentiment = data.sentiment_score;\n\n// Level classifiers\nconst growthLevel = growth >= 0.25 ? 'HIGH' : growth >= 0.10 ? 'MEDIUM' : 'LOW';\nconst sentimentLevel = sentiment >= 0.65 ? 'HIGH' : sentiment >= 0.40 ? 'MEDIUM' : 'LOW';\n\n// 3Γ—3 Signal Matrix β€” each cell is a full decision profile\nconst signalMatrix = {\n HIGH_HIGH: { signal:'STRONG_BUY', confidence:0.88, risk:'LOW', short_term:'+15 to +25%', long_term:'+60 to +120%', description:'Both fundamentals and market conviction are aligned β€” ideal setup.' },\n HIGH_MEDIUM: { signal:'BUY', confidence:0.72, risk:'LOW_MEDIUM', short_term:'+8 to +15%', long_term:'+40 to +80%', description:'Strong fundamentals; market is starting to recognise the value.' },\n HIGH_LOW: { signal:'FUNDAMENTAL_BUY', confidence:0.61, risk:'MEDIUM', short_term:'-5 to +5%', long_term:'+30 to +70%', description:'Market undervaluing strong fundamentals β€” contrarian value play.' },\n MEDIUM_HIGH: { signal:'HOLD_WATCH', confidence:0.55, risk:'MEDIUM', short_term:'+5 to +12%', long_term:'+15 to +35%', description:'Sentiment slightly ahead of fundamentals β€” monitor for convergence.' },\n MEDIUM_MEDIUM:{ signal:'HOLD', confidence:0.65, risk:'MEDIUM', short_term:'0 to +8%', long_term:'+10 to +25%', description:'Balanced but lacks a strong catalyst either way.' },\n MEDIUM_LOW: { signal:'UNDERPERFORM', confidence:0.60, risk:'MEDIUM_HIGH', short_term:'-5 to +5%', long_term:'+5 to +15%', description:'Adequate fundamentals dragged down by weak sentiment.' },\n LOW_HIGH: { signal:'BUBBLE_RISK', confidence:0.75, risk:'HIGH', short_term:'+5 to +15%', long_term:'-20 to 0%', description:'DANGER: Sentiment massively disconnected from weak fundamentals.' },\n LOW_MEDIUM: { signal:'SELL', confidence:0.68, risk:'HIGH', short_term:'-5 to -15%', long_term:'-10 to +10%', description:'Deteriorating fundamentals with fading sentiment support.' },\n LOW_LOW: { signal:'STRONG_SELL', confidence:0.82, risk:'VERY_HIGH', short_term:'-15 to -30%', long_term:'-30 to -10%', description:'Both fundamentals and market sentiment fully bearish.' }\n};\n\nconst matrixKey = `${growthLevel}_${sentimentLevel}`;\nconst result = signalMatrix[matrixKey];\n\n// Conflict & Bubble Detection\nconst conflictDetected = (growthLevel === 'HIGH' && sentimentLevel === 'LOW') ||\n (growthLevel === 'LOW' && sentimentLevel === 'HIGH');\n\nconst bubbleRiskScore =\n (growthLevel === 'LOW' && sentimentLevel === 'HIGH') ? 0.85 :\n (growthLevel === 'MEDIUM' && sentimentLevel === 'HIGH') ? 0.48 :\n (growthLevel === 'LOW' && sentimentLevel === 'MEDIUM') ? 0.35 : 0.12;\n\nconst correctionProbability =\n bubbleRiskScore > 0.70 ? 'HIGH (>65%)' :\n bubbleRiskScore > 0.40 ? 'MEDIUM (30–65%)' : 'LOW (<30%)';\n\n// Key Insight β€” human-readable narrative\nlet keyInsight;\nif (growthLevel === 'HIGH' && sentimentLevel === 'LOW') {\n keyInsight = `⚠️ CONFLICT DETECTED: Strong ${Math.round(growth*100)}% revenue growth is being ignored by the market (sentiment ${(sentiment*100).toFixed(0)}%). Classic value opportunity β€” patience required.`;\n} else if (growthLevel === 'LOW' && sentimentLevel === 'HIGH') {\n keyInsight = `🚨 BUBBLE RISK: High sentiment (${(sentiment*100).toFixed(0)}%) is NOT supported by ${Math.round(growth*100)}% growth. Correction probability is ${correctionProbability}.`;\n} else {\n keyInsight = `${growthLevel} growth + ${sentimentLevel} sentiment β†’ ${result.signal}. ${result.description}`;\n}\n\nreturn [{ json: {\n ...data,\n sim1_growth_sentiment: {\n simulation: 'GROWTH_VS_SENTIMENT',\n inputs: { growth_rate: growth, sentiment_score: sentiment, growth_level: growthLevel, sentiment_level: sentimentLevel },\n output: result,\n conflict_detected: conflictDetected,\n bubble_risk_score: bubbleRiskScore,\n correction_probability: correctionProbability,\n key_insight: keyInsight\n }\n} }];" }, "id": "b9e5fe29-be12-498b-b0db-aa96f8c066c4", "name": "πŸ“ˆ Sim 1 β€” Growth vs Sentiment Tradeoff", "type": "n8n-nodes-base.code", "typeVersion": 2, "position": [ -1872, 96 ] }, { "parameters": { "jsCode": "// ─── SIMULATION 2: Risk Appetite Simulator ───────────────────────────────────\n// Core Principle: Same stock β†’ different decisions for different investor types\nconst data = $input.first().json;\nconst riskLevel = data.risk_appetite;\nconst growth = data.revenue_growth;\nconst sentiment = data.sentiment_score;\nconst pe = data.pe_ratio;\nconst de = data.debt_equity;\n\n// Full investor profile definitions\nconst profiles = {\n low: {\n name: 'Conservative', icon: 'πŸ”΅',\n growth_threshold: 0.15,\n sentiment_threshold: 0.60,\n pe_max: 25,\n de_max: 0.50,\n max_position_pct: 5,\n stop_loss_pct: 5,\n take_profit_pct: 12,\n preferred_horizon: '3–5 years',\n max_drawdown: '10%',\n description: 'Capital preservation is paramount. Prefers stable, established companies with low volatility and dividends.'\n },\n medium: {\n name: 'Balanced', icon: '🟒',\n growth_threshold: 0.10,\n sentiment_threshold: 0.45,\n pe_max: 40,\n de_max: 0.80,\n max_position_pct: 10,\n stop_loss_pct: 10,\n take_profit_pct: 25,\n preferred_horizon: '1–3 years',\n max_drawdown: '20%',\n description: 'Balanced growth and protection. Accepts moderate volatility in exchange for higher long-term returns.'\n },\n high: {\n name: 'Aggressive', icon: '🟠',\n growth_threshold: 0.05,\n sentiment_threshold: 0.35,\n pe_max: 80,\n de_max: 1.50,\n max_position_pct: 20,\n stop_loss_pct: 20,\n take_profit_pct: 60,\n preferred_horizon: '6–18 months',\n max_drawdown: '40%',\n description: 'Maximum growth orientation. High volatility tolerance. Seeks asymmetric upside opportunities.'\n }\n};\n\nconst profile = profiles[riskLevel];\n\n// Evaluate this stock against the selected profile β€” 4 weighted criteria\nconst criteria = [\n { name: 'Revenue Growth', pass: growth >= profile.growth_threshold, actual: `${Math.round(growth*100)}%`, required: `β‰₯ ${Math.round(profile.growth_threshold*100)}%`, weight: 0.30 },\n { name: 'Market Sentiment',pass: sentiment >= profile.sentiment_threshold, actual: sentiment.toFixed(2), required: `β‰₯ ${profile.sentiment_threshold.toFixed(2)}`, weight: 0.25 },\n { name: 'P/E Ratio', pass: pe <= profile.pe_max, actual: pe.toFixed(1), required: `≀ ${profile.pe_max}`, weight: 0.25 },\n { name: 'Debt / Equity', pass: de <= profile.de_max, actual: de.toFixed(2), required: `≀ ${profile.de_max}`, weight: 0.20 }\n];\n\nconst weightedPassRate = criteria.reduce((sum, c) => sum + (c.pass ? c.weight : 0), 0);\nconst passCount = criteria.filter(c => c.pass).length;\n\nconst recommendation =\n weightedPassRate >= 0.80 ? 'ENTER_FULL_POSITION' :\n weightedPassRate >= 0.55 ? 'ENTER_PARTIAL_POSITION' :\n weightedPassRate >= 0.30 ? 'WATCHLIST_ONLY' : 'AVOID';\n\n// Cross-profile comparison β€” how does the SAME stock look to each investor type?\nconst crossProfileComparison = Object.entries(profiles).map(([level, p]) => {\n const criteriaScores = [\n { pass: growth >= p.growth_threshold, weight: 0.30 },\n { pass: sentiment >= p.sentiment_threshold, weight: 0.25 },\n { pass: pe <= p.pe_max, weight: 0.25 },\n { pass: de <= p.de_max, weight: 0.20 }\n ];\n const wpr = criteriaScores.reduce((s, c) => s + (c.pass ? c.weight : 0), 0);\n const rec = wpr >= 0.80 ? 'ENTER' : wpr >= 0.55 ? 'PARTIAL' : wpr >= 0.30 ? 'WATCH' : 'AVOID';\n return {\n risk_level: level,\n investor_type: p.name,\n fit_score_pct: Math.round(wpr * 100),\n recommendation: rec,\n max_position: `${p.max_position_pct}%`,\n stop_loss: `${p.stop_loss_pct}%`,\n is_selected: level === riskLevel\n };\n});\n\nconst keyInsight = `${profile.icon} ${profile.name} investor β†’ ${recommendation} | ${Math.round(weightedPassRate*100)}% criteria met (${passCount}/4 passed) | Max position: ${profile.max_position_pct}% | Stop-loss: ${profile.stop_loss_pct}%`;\n\nreturn [{ json: {\n ...data,\n sim2_risk_appetite: {\n simulation: 'RISK_APPETITE',\n inputs: { risk_appetite: riskLevel, pe_ratio: pe, debt_equity: de, revenue_growth: growth, sentiment_score: sentiment },\n profile,\n criteria_evaluation: criteria,\n weighted_pass_rate: Math.round(weightedPassRate * 100) / 100,\n pass_count: `${passCount}/4`,\n recommendation,\n cross_profile_comparison: crossProfileComparison,\n key_insight: keyInsight\n }\n} }];" }, "id": "8f0dafcd-595b-41e8-8be9-421e85d80bd6", "name": "βš–οΈ Sim 2 β€” Risk Appetite Simulator", "type": "n8n-nodes-base.code", "typeVersion": 2, "position": [ -1632, 96 ] }, { "parameters": { "jsCode": "// ─── SIMULATION 3: Threshold Break Simulator ─────────────────────────────────\n// Core Principle: At what exact point does the recommendation CHANGE?\nconst data = $input.first().json;\nconst currentGrowth = data.revenue_growth;\nconst currentSentiment = data.sentiment_score;\n\n// Shared signal classifier (mirrors Sim1 logic)\nfunction classify(g, s) {\n const gL = g >= 0.25 ? 'HIGH' : g >= 0.10 ? 'MEDIUM' : 'LOW';\n const sL = s >= 0.65 ? 'HIGH' : s >= 0.40 ? 'MEDIUM' : 'LOW';\n const m = {\n HIGH_HIGH:'STRONG_BUY', HIGH_MEDIUM:'BUY', HIGH_LOW:'FUNDAMENTAL_BUY',\n MEDIUM_HIGH:'HOLD_WATCH', MEDIUM_MEDIUM:'HOLD', MEDIUM_LOW:'UNDERPERFORM',\n LOW_HIGH:'BUBBLE_RISK', LOW_MEDIUM:'SELL', LOW_LOW:'STRONG_SELL'\n };\n return m[`${gL}_${sL}`] || 'HOLD';\n}\n\nconst baseRecommendation = classify(currentGrowth, currentSentiment);\n\n// Helper: find breakpoints by sweeping a variable\nfunction findBreakpoints(sweepVar, fixedGrowth, fixedSentiment, min, max, step) {\n const sweep = [];\n for (let v = min; v <= max + 0.0001; v = Math.round((v + step) * 1000) / 1000) {\n const g = sweepVar === 'growth' ? v : fixedGrowth;\n const s = sweepVar === 'sentiment' ? v : fixedSentiment;\n sweep.push({ value: v, recommendation: classify(g, s) });\n }\n const breaks = [];\n for (let i = 1; i < sweep.length; i++) {\n if (sweep[i].recommendation !== sweep[i-1].recommendation) {\n const threshold = Math.round(((sweep[i-1].value + sweep[i].value) / 2) * 1000) / 1000;\n const currentValue = sweepVar === 'growth' ? fixedGrowth : fixedSentiment;\n const distFromCurrent = Math.abs(threshold - currentValue);\n breaks.push({\n transition: `${sweep[i-1].recommendation} β†’ ${sweep[i].recommendation}`,\n threshold_approx: threshold,\n direction: sweep[i].value > currentValue ? 'MUST_INCREASE_TO_TRIGGER' : 'BREAKS_IF_DECREASES',\n distance_from_current: Math.round(distFromCurrent * 1000) / 1000,\n is_nearest: false\n });\n }\n }\n // Mark nearest breakpoint\n if (breaks.length > 0) {\n const nearestIdx = breaks.reduce((ni, b, i, arr) =>\n b.distance_from_current < arr[ni].distance_from_current ? i : ni, 0);\n breaks[nearestIdx].is_nearest = true;\n }\n return breaks;\n}\n\n// Sweep sentiment (0.00 β†’ 1.00, step 0.05)\nconst sentimentBreakpoints = findBreakpoints('sentiment', currentGrowth, currentSentiment, 0.00, 1.00, 0.05);\n\n// Sweep growth (-0.10 β†’ 0.60, step 0.05)\nconst growthBreakpoints = findBreakpoints('growth', currentGrowth, currentSentiment, -0.10, 0.60, 0.05);\n\n// Sensitivity scoring: how many breakpoints lie within Β±0.15 of current values\nconst sentimentBreaksNearby = sentimentBreakpoints.filter(b => b.distance_from_current <= 0.15).length;\nconst growthBreaksNearby = growthBreakpoints.filter(b => b.distance_from_current <= 0.10).length;\nconst primaryDriver = sentimentBreaksNearby >= growthBreaksNearby ? 'SENTIMENT' : 'GROWTH';\n\n// Stability score: distance from nearest breakpoints (1.0 = very stable, 0.0 = at a breakpoint)\nconst minSentDist = sentimentBreakpoints.length > 0 ? Math.min(...sentimentBreakpoints.map(b => b.distance_from_current)) : 0.50;\nconst minGrowthDist = growthBreakpoints.length > 0 ? Math.min(...growthBreakpoints.map(b => b.distance_from_current)) : 0.50;\nconst stabilityScore = Math.round(Math.min(1, (minSentDist * 2.0 + minGrowthDist * 3.0) / 2.5) * 100) / 100;\n\n// Two-way stress tests: what does Β±10% change in each variable cause?\nconst stressTests = [\n { scenario: 'Sentiment βˆ’10%', g: currentGrowth, s: Math.max(0, currentSentiment - 0.10), result: classify(currentGrowth, Math.max(0, currentSentiment - 0.10)) },\n { scenario: 'Sentiment +10%', g: currentGrowth, s: Math.min(1, currentSentiment + 0.10), result: classify(currentGrowth, Math.min(1, currentSentiment + 0.10)) },\n { scenario: 'Growth βˆ’10pp', g: currentGrowth - 0.10, s: currentSentiment, result: classify(currentGrowth - 0.10, currentSentiment) },\n { scenario: 'Growth +10pp', g: currentGrowth + 0.10, s: currentSentiment, result: classify(currentGrowth + 0.10, currentSentiment) },\n { scenario: 'Both βˆ’10%', g: currentGrowth - 0.10, s: Math.max(0, currentSentiment - 0.10), result: classify(currentGrowth - 0.10, Math.max(0, currentSentiment - 0.10)) },\n { scenario: 'Both +10%', g: currentGrowth + 0.10, s: Math.min(1, currentSentiment + 0.10), result: classify(currentGrowth + 0.10, Math.min(1, currentSentiment + 0.10)) }\n].map(t => ({\n scenario: t.scenario,\n result: t.result,\n changed: t.result !== baseRecommendation,\n delta: t.result !== baseRecommendation ? `${baseRecommendation} β†’ ${t.result}` : 'No change'\n}));\n\n// Number of stress tests that triggered a recommendation change\nconst stressBreachCount = stressTests.filter(t => t.changed).length;\nconst stressResiliency = stressBreachCount === 0 ? 'HIGH' : stressBreachCount <= 2 ? 'MEDIUM' : 'LOW';\n\nconst nearestSentBreak = sentimentBreakpoints.find(b => b.is_nearest);\nconst nearestGrowthBreak = growthBreakpoints.find(b => b.is_nearest);\n\nlet keyInsight;\nif (nearestSentBreak && nearestSentBreak.distance_from_current < 0.10) {\n keyInsight = `🚨 HIGH SENSITIVITY: Decision is ${(nearestSentBreak.distance_from_current * 100).toFixed(0)} sentiment-points from a change (${nearestSentBreak.transition}). Primary driver: ${primaryDriver}.`;\n} else if (stabilityScore > 0.7) {\n keyInsight = `βœ… STABLE: Current recommendation (${baseRecommendation}) is robust. Nearest trigger is ${(minSentDist * 100).toFixed(0)} sentiment-points away. Primary driver: ${primaryDriver}.`;\n} else {\n keyInsight = `⚠️ MODERATE SENSITIVITY: ${primaryDriver} is the primary driver. ${nearestSentBreak ? `Sentiment threshold: ${nearestSentBreak.transition} at ~${nearestSentBreak.threshold_approx}` : ''}. Stability: ${stabilityScore}.`;\n}\n\nreturn [{ json: {\n ...data,\n sim3_threshold_break: {\n simulation: 'THRESHOLD_BREAK',\n current_state: { recommendation: baseRecommendation, growth: currentGrowth, sentiment: currentSentiment },\n sentiment_breakpoints: sentimentBreakpoints,\n growth_breakpoints: growthBreakpoints,\n stress_tests: stressTests,\n sensitivity: {\n sentiment_breaks_nearby: sentimentBreaksNearby,\n growth_breaks_nearby: growthBreaksNearby,\n primary_driver: primaryDriver,\n stress_resiliency: stressResiliency,\n stress_breach_count: `${stressBreachCount}/6`\n },\n stability_score: stabilityScore,\n key_insight: keyInsight\n }\n} }];" }, "id": "6ed385b8-211e-4c6b-af7f-dde724144a36", "name": "🎯 Sim 3 β€” Threshold Break Simulator", "type": "n8n-nodes-base.code", "typeVersion": 2, "position": [ -1392, 96 ] }, { "parameters": { "jsCode": "// ─── AGGREGATE ENGINE: Combine all 3 simulations into a final verdict ─────────\nconst data = $input.first().json;\nconst sim1 = data.sim1_growth_sentiment;\nconst sim2 = data.sim2_risk_appetite;\nconst sim3 = data.sim3_threshold_break;\n\n// Convert signals to numeric scores (1–5 scale)\nconst signalScores = {\n 'STRONG_BUY':5.0, 'BUY':4.2, 'FUNDAMENTAL_BUY':3.8,\n 'HOLD_WATCH':3.2, 'HOLD':3.0, 'UNDERPERFORM':2.2,\n 'BUBBLE_RISK':1.8, 'SELL':1.5, 'STRONG_SELL':1.0\n};\nconst recScores = {\n 'ENTER_FULL_POSITION':5.0, 'ENTER_PARTIAL_POSITION':3.8,\n 'WATCHLIST_ONLY':2.5, 'AVOID':1.0\n};\n\n// Weighted scoring: Sim1 most important (market signal), Sim2 next (fit), Sim3 as risk modifier\nconst s1Raw = (signalScores[sim1.output.signal] || 3.0) * sim1.output.confidence;\nconst s2Raw = recScores[sim2.recommendation] || 3.0;\nconst s3Raw = sim3.stability_score * 5.0;\n\nconst weights = { sim1: 0.45, sim2: 0.35, sim3: 0.20 };\nconst rawScore = (s1Raw * weights.sim1) + (s2Raw * weights.sim2) + (s3Raw * weights.sim3);\nconst score = Math.round(Math.min(5.0, Math.max(1.0, rawScore)) * 100) / 100;\n\nconst finalRecommendation =\n score >= 4.2 ? 'STRONG_BUY' :\n score >= 3.6 ? 'BUY' :\n score >= 2.8 ? 'HOLD' :\n score >= 1.8 ? 'SELL' : 'STRONG_SELL';\n\n// ── Probabilistic Scenario Engine ────────────────────────────────────────────\nconst norm = score / 5.0;\nconst thorizonMult = data.time_horizon === '5year' ? 3.0 : data.time_horizon === '3month' ? 0.25 : 1.0;\nconst baseReturn = (score - 3.0) * 18; // % return expected per year at base\n\nconst bestCaseProb = Math.round(Math.min(65, Math.max(5, norm * 60 + 5)));\nconst worstCaseProb = Math.round(Math.min(65, Math.max(5, (1 - norm) * 55 + 5)));\nconst expectedProb = 100 - bestCaseProb - worstCaseProb;\n\nconst scenarios = {\n best_case: {\n probability: bestCaseProb,\n label: 'All tailwinds materialise',\n return_estimate:`+${Math.round((baseReturn + 28) * thorizonMult)}%`,\n conditions: ['Revenue growth beats estimates', 'Positive sentiment momentum', 'Macro conditions remain favourable', 'No sector disruptions']\n },\n expected_case: {\n probability: expectedProb,\n label: 'Base case β€” current trajectory continues',\n return_estimate:`${baseReturn >= 0 ? '+' : ''}${Math.round(baseReturn * thorizonMult)}%`,\n conditions: ['Current growth rate maintained', 'Sentiment stable', 'Normal market conditions', 'No major earnings surprises']\n },\n worst_case: {\n probability: worstCaseProb,\n label: 'Headwinds emerge',\n return_estimate:`-${Math.round((Math.abs(baseReturn) + 22) * thorizonMult)}%`,\n conditions: ['Sentiment deteriorates sharply', 'Revenue misses estimates', 'Macro headwinds', 'Sector rotation away']\n }\n};\n\n// ── Risk Flag Engine ──────────────────────────────────────────────────────────\nconst riskFlags = [];\nif (sim1.conflict_detected) riskFlags.push({ severity:'MEDIUM', flag:'Growth–Sentiment conflict detected', action:'Consider scaling in gradually' });\nif (sim1.bubble_risk_score > 0.60) riskFlags.push({ severity:'HIGH', flag:`Bubble risk score: ${sim1.bubble_risk_score}`, action:'Avoid new positions; set tight stop-loss' });\nif (sim3.stability_score < 0.40) riskFlags.push({ severity:'HIGH', flag:'Recommendation near multiple breakpoints', action:'Monitor daily; do not hold large position' });\nif (sim3.sensitivity.sentiment_breaks_nearby > 2) riskFlags.push({ severity:'MEDIUM', flag:'Highly sentiment-sensitive β€” 3+ triggers nearby', action:'Watch social/news sentiment closely' });\nif (data.pe_ratio > 60 && data.revenue_growth < 0.20) riskFlags.push({ severity:'HIGH', flag:`P/E ${data.pe_ratio}x not justified by ${Math.round(data.revenue_growth*100)}% growth`, action:'Valuations stretched β€” wait for pullback' });\nif (data.debt_equity > 1.0) riskFlags.push({ severity:'MEDIUM', flag:`Elevated D/E ratio: ${data.debt_equity}`, action:'Monitor debt levels and cash flow' });\nif (parseInt(sim3.sensitivity.stress_breach_count) > 4) riskFlags.push({ severity:'HIGH', flag:'Low stress resiliency β€” 5+ scenarios trigger change', action:'Only trade with defined risk limits' });\n\n// ── Confidence Score ──────────────────────────────────────────────────────────\nconst confidenceInputs = [sim1.output.confidence, sim2.weighted_pass_rate, sim3.stability_score];\nconst overallConfidence = Math.round(confidenceInputs.reduce((a, b) => a + b, 0) / 3 * 100);\n\nreturn [{ json: {\n ...data,\n aggregate: {\n final_recommendation: finalRecommendation,\n score,\n score_label: `${score}/5.0`,\n confidence_pct: overallConfidence,\n simulation_weights: weights,\n component_scores: { sim1_weighted: Math.round(s1Raw * weights.sim1 * 100)/100, sim2_weighted: Math.round(s2Raw * weights.sim2 * 100)/100, sim3_weighted: Math.round(s3Raw * weights.sim3 * 100)/100 },\n scenarios,\n risk_flags: riskFlags,\n risk_flag_count: riskFlags.length,\n primary_risk_driver: sim3.sensitivity.primary_driver,\n key_signals: [\n sim1.key_insight,\n sim2.key_insight,\n sim3.key_insight\n ].filter(Boolean)\n }\n} }];" }, "id": "be1bfc4e-7109-41bf-9030-c4e13422af12", "name": "🧠 Aggregate & Score All Simulations", "type": "n8n-nodes-base.code", "typeVersion": 2, "position": [ -1152, 96 ] }, { "parameters": { "method": "POST", "url": "https://api.groq.com/openai/v1/chat/completions", "authentication": "predefinedCredentialType", "nodeCredentialType": "groqApi", "sendBody": true, "specifyBody": "json", "jsonBody": "={{ JSON.stringify({ model: 'llama-3.3-70b-versatile', max_tokens: 1500, temperature: 0.3, messages: [{ role: 'system', content: 'You are Mycroft, an expert AI investment analyst. You always respond with valid, parseable JSON only. No markdown code blocks, no preamble, no trailing text outside the JSON.' }, { role: 'user', content: $json.ai_prompt }] }) }}", "options": { "response": { "response": {} }, "timeout": 30000 } }, "id": "8ef16a58-8fa0-4c7d-8194-c7af8d999824", "name": "πŸ€– Groq β€” Generate Investment Insights", "type": "n8n-nodes-base.httpRequest", "typeVersion": 4.2, "position": [ -672, 96 ] }, { "parameters": { "jsCode": "// ─── RESPONSE PARSER: Extract structured insights from Groq (OpenAI format) ───\nconst apiResponse = $input.first().json;\n// Retrieve full simulation data from the last code node before the HTTP call\nconst prevData = $('πŸ“ Build AI Prompt').first().json;\n\n// Groq uses OpenAI-compatible format: choices[0].message.content\nlet aiText = '';\nif (apiResponse.choices && Array.isArray(apiResponse.choices) && apiResponse.choices.length > 0) {\n aiText = apiResponse.choices[0]?.message?.content || '';\n} else if (apiResponse.error) {\n // Groq error response\n throw new Error(`Groq API Error: ${apiResponse.error.message || JSON.stringify(apiResponse.error)}`);\n}\n\n// Parse JSON β€” strip potential markdown fences if the model added them\nlet aiInsights = {};\ntry {\n const cleaned = aiText\n .replace(/^```json\\s*/i, '')\n .replace(/^```\\s*/i, '')\n .replace(/```\\s*$/, '')\n .trim();\n aiInsights = JSON.parse(cleaned);\n} catch (parseErr) {\n // Fallback: construct a basic insights object from raw text\n console.error('Groq JSON parse failed:', parseErr.message);\n aiInsights = {\n executive_summary: aiText.substring(0, 300) || 'AI analysis unavailable β€” review quantitative data.',\n bull_case: 'See simulation results for fundamental signals.',\n bear_case: 'See risk flags for downside scenarios.',\n key_risk_narrative: `Primary risk driver: ${prevData.aggregate.primary_risk_driver || 'SENTIMENT'}.`,\n opportunity_narrative:'Review Sim 1 growth vs sentiment analysis.',\n action_plan: {\n immediate: `${prevData.aggregate.final_recommendation} based on quantitative analysis`,\n monitor_weekly: 'Revenue growth rate and sentiment score',\n exit_trigger: 'If score drops below 2.8, revisit recommendation',\n position_sizing_note:`See risk appetite simulation for position guidance`\n },\n mycroft_verdict: `${prevData.aggregate.final_recommendation} β€” score ${prevData.aggregate.score_label} with ${prevData.aggregate.confidence_pct}% confidence.`,\n confidence_rationale: `Confidence based on ${prevData.aggregate.confidence_pct}% pass rate across 3 simulation engines.`\n };\n}\n\n// Log Groq token usage for cost tracking (Groq is free but has rate limits)\nconst usage = apiResponse.usage || {};\nconsole.log(`Groq tokens β€” prompt: ${usage.prompt_tokens || 0}, completion: ${usage.completion_tokens || 0}, total: ${usage.total_tokens || 0}`);\n\nreturn [{ json: { ...prevData, ai_insights: aiInsights, ai_raw_text: aiText, groq_usage: usage } }];" }, "id": "ef673e08-7919-484d-bf05-b9ae48f3b70f", "name": "πŸ”„ Parse Groq Response", "type": "n8n-nodes-base.code", "typeVersion": 2, "position": [ -432, 96 ] }, { "parameters": { "jsCode": "// ─── DB PREP: Sanitise and structure data for PostgreSQL insertion ─────────────\n// NOTE: For production use, replace string interpolation with parameterised queries.\nconst d = $input.first().json;\n\n// Safe string escape for SQL (MVP-grade β€” use pgp or parameterised queries in prod)\nconst esc = (v) => String(v ?? '').replace(/'/g, \"''\");\n\nconst inputParamsJson = esc(JSON.stringify({\n revenue_growth: d.revenue_growth,\n sentiment_score: d.sentiment_score,\n pe_ratio: d.pe_ratio,\n debt_equity: d.debt_equity,\n current_price: d.current_price,\n market_cap_billions: d.market_cap_billions\n}));\n\nconst simResultsJson = esc(JSON.stringify({\n sim1: d.sim1_growth_sentiment,\n sim2: d.sim2_risk_appetite,\n sim3: d.sim3_threshold_break,\n aggregate: d.aggregate\n}));\n\nconst aiInsightsJson = esc(JSON.stringify(d.ai_insights || {}));\n\nconst dbQuery = `\nINSERT INTO mycroft_whatif_simulations (\n simulation_id,\n company,\n ticker,\n risk_appetite,\n time_horizon,\n final_recommendation,\n confidence_pct,\n aggregate_score,\n bubble_risk_score,\n stability_score,\n input_params,\n simulation_results,\n ai_insights,\n ai_verdict\n) VALUES (\n '${esc(d.simulation_id)}',\n '${esc(d.company)}',\n '${esc(d.ticker)}',\n '${esc(d.risk_appetite)}',\n '${esc(d.time_horizon)}',\n '${esc(d.aggregate.final_recommendation)}',\n ${d.aggregate.confidence_pct},\n ${d.aggregate.score},\n ${d.sim1_growth_sentiment.bubble_risk_score},\n ${d.sim3_threshold_break.stability_score},\n '${inputParamsJson}'::jsonb,\n '${simResultsJson}'::jsonb,\n '${aiInsightsJson}'::jsonb,\n '${esc(d.ai_insights?.mycroft_verdict || '')}'\n)\nON CONFLICT (simulation_id) DO UPDATE\n SET ai_verdict = EXCLUDED.ai_verdict,\n final_recommendation = EXCLUDED.final_recommendation\nRETURNING id, simulation_id;\n`.trim();\n\nreturn [{ json: { ...d, db_query: dbQuery } }];" }, "id": "b7a1860a-f042-4a83-bc43-cd2a985d27e9", "name": "πŸ—„οΈ Prepare DB Record", "type": "n8n-nodes-base.code", "typeVersion": 2, "position": [ -192, 96 ] }, { "parameters": { "operation": "executeQuery", "query": "={{ $json.db_query }}", "options": { "queryBatching": "single" } }, "id": "f5a82825-d117-4fbe-9a9a-44dc46b26059", "name": "πŸ’Ύ Save to PostgreSQL", "type": "n8n-nodes-base.postgres", "typeVersion": 2.5, "position": [ 48, 96 ] }, { "parameters": { "jsCode": "// ─── RESPONSE BUILDER: Assemble the final structured API response ─────────────\n// PostgreSQL node returns only the RETURNING row β€” we need to pull full data back\nconst fullData = $('πŸ”„ Parse Groq Response').first().json;\nconst dbResult = $input.first().json; // Has { id, simulation_id } from RETURNING\n\nconst d = fullData;\nconst agg = d.aggregate;\nconst s1 = d.sim1_growth_sentiment;\nconst s2 = d.sim2_risk_appetite;\nconst s3 = d.sim3_threshold_break;\nconst ai = d.ai_insights;\n\n// Recommendation icons\nconst recEmoji = {\n 'STRONG_BUY':'πŸš€', 'BUY':'πŸ“ˆ', 'FUNDAMENTAL_BUY':'πŸ’Ž',\n 'HOLD_WATCH':'πŸ‘€', 'HOLD':'βš–οΈ', 'UNDERPERFORM':'😐',\n 'BUBBLE_RISK':'⚠️', 'SELL':'πŸ“‰', 'STRONG_SELL':'🚨'\n};\n\nconst processingMs = Date.now() - new Date(d.timestamp).getTime();\n\nconst response = {\n status: 'SUCCESS',\n simulation_id: d.simulation_id,\n db_record_id: dbResult.id || null,\n timestamp: d.timestamp,\n processed_in_ms: processingMs,\n\n // ── Company ──────────────────────────────────────────────────────────────\n company: {\n name: d.company,\n ticker: d.ticker,\n current_price: `$${d.current_price}`,\n market_cap: `$${d.market_cap_billions}B`\n },\n\n // ── Final Verdict ─────────────────────────────────────────────────────────\n verdict: {\n emoji: recEmoji[agg.final_recommendation] || '❓',\n recommendation: agg.final_recommendation,\n score: agg.score_label,\n confidence: `${agg.confidence_pct}%`,\n mycroft_says: ai.mycroft_verdict || `${agg.final_recommendation} β€” based on 3-simulation analysis`\n },\n\n // ── AI Narrative ──────────────────────────────────────────────────────────\n analysis: {\n executive_summary: ai.executive_summary,\n bull_case: ai.bull_case,\n bear_case: ai.bear_case,\n opportunity: ai.opportunity_narrative,\n key_risk: ai.key_risk_narrative,\n confidence_rationale: ai.confidence_rationale\n },\n\n // ── Action Plan ───────────────────────────────────────────────────────────\n action_plan: {\n immediate: ai.action_plan?.immediate || `Review ${agg.final_recommendation} signal`,\n monitor_weekly: ai.action_plan?.monitor_weekly || `${agg.primary_risk_driver} score`,\n exit_trigger: ai.action_plan?.exit_trigger || 'Significant deterioration in primary driver',\n position_sizing: ai.action_plan?.position_sizing_note || `Max ${s2.profile.max_position_pct}% of portfolio (${s2.profile.name} profile)`\n },\n\n // ── Simulation Detail ─────────────────────────────────────────────────────\n simulations: {\n sim1_growth_vs_sentiment: {\n label: 'πŸ“ˆ Growth vs Sentiment Tradeoff',\n signal: s1.output.signal,\n confidence: `${Math.round(s1.output.confidence * 100)}%`,\n conflict: s1.conflict_detected,\n bubble_risk: s1.bubble_risk_score,\n expected_returns: { short_term: s1.output.short_term, long_term: s1.output.long_term },\n insight: s1.key_insight\n },\n sim2_risk_appetite: {\n label: 'βš–οΈ Risk Appetite Simulator',\n investor_type: s2.profile.name,\n recommendation: s2.recommendation,\n criteria_passed: s2.pass_count,\n weighted_fit: `${Math.round(s2.weighted_pass_rate * 100)}%`,\n position_limit: `${s2.profile.max_position_pct}% of portfolio`,\n stop_loss: `${s2.profile.stop_loss_pct}%`,\n take_profit: `${s2.profile.take_profit_pct}%`,\n cross_profile: s2.cross_profile_comparison,\n insight: s2.key_insight\n },\n sim3_threshold_break: {\n label: '🎯 Threshold Break Simulator',\n current_signal: s3.current_state.recommendation,\n stability_score: s3.stability_score,\n primary_driver: s3.sensitivity.primary_driver,\n stress_resiliency: s3.sensitivity.stress_resiliency,\n stress_tests: s3.stress_tests,\n critical_thresholds: [\n ...s3.sentiment_breakpoints.filter(b => b.is_nearest).map(b => ({ type:'SENTIMENT', ...b })),\n ...s3.growth_breakpoints.filter(b => b.is_nearest).map(b => ({ type:'GROWTH', ...b }))\n ],\n insight: s3.key_insight\n }\n },\n\n // ── Scenarios ─────────────────────────────────────────────────────────────\n scenarios: {\n best_case: agg.scenarios.best_case,\n expected_case: agg.scenarios.expected_case,\n worst_case: agg.scenarios.worst_case\n },\n\n // ── Risk Summary ──────────────────────────────────────────────────────────\n risk_summary: {\n flag_count: agg.risk_flag_count,\n overall_risk_level: agg.risk_flags.some(f => f.severity === 'HIGH') ? 'HIGH' : agg.risk_flags.some(f => f.severity === 'MEDIUM') ? 'MEDIUM' : 'LOW',\n flags: agg.risk_flags,\n primary_driver: agg.primary_risk_driver\n },\n\n // ── Metadata ──────────────────────────────────────────────────────────────\n metadata: {\n model: 'Mycroft-WhatIf-MVP-v1.0',\n simulations_run: 3,\n ai_model: 'groq/llama-3.3-70b-versatile',\n processing_ms: processingMs,\n investor_profile: { risk: d.risk_appetite, horizon: d.time_horizon }\n }\n};\n\nreturn [{ json: response }];" }, "id": "cbd4aa15-0aa9-426a-a145-c818eba4833a", "name": "πŸ“Š Build Final Response", "type": "n8n-nodes-base.code", "typeVersion": 2, "position": [ 288, 96 ] }, { "parameters": { "respondWith": "json", "responseBody": "={{ $json }}", "options": { "responseCode": 200, "responseHeaders": { "entries": [ { "name": "Content-Type", "value": "application/json" }, { "name": "X-Mycroft-Version", "value": "1.0.0" }, { "name": "X-Simulation-ID", "value": "={{ $json.simulation_id }}" }, { "name": "X-Recommendation", "value": "={{ $json.verdict.recommendation }}" }, { "name": "X-Confidence", "value": "={{ $json.verdict.confidence }}" } ] } } }, "id": "42466b0e-2b18-4bfd-b117-df5f9136cc17", "name": "βœ… Respond to Webhook", "type": "n8n-nodes-base.respondToWebhook", "typeVersion": 1.1, "position": [ 528, 96 ] }, { "parameters": { "jsCode": "// ─── PROMPT ENGINEER: Build structured analysis prompt for Claude ─────────────\nconst d = $input.first().json;\nconst agg = d.aggregate;\nconst s1 = d.sim1_growth_sentiment;\nconst s2 = d.sim2_risk_appetite;\nconst s3 = d.sim3_threshold_break;\n\nconst prompt = `You are Mycroft, an expert AI investment analyst from the Mycroft open-source project.\nAnalyse the following What-If simulation output and return a structured JSON investment report.\n\n═══════════════════════════════════════════\nCOMPANY PROFILE\n═══════════════════════════════════════════\nCompany: ${d.company} (${d.ticker})\nCurrent Price: $${d.current_price}\nRevenue Growth: ${Math.round(d.revenue_growth * 100)}% YoY\nSentiment Score: ${d.sentiment_score}/1.0\nP/E Ratio: ${d.pe_ratio}x\nDebt/Equity: ${d.debt_equity}\nMarket Cap: $${d.market_cap_billions}B\nRisk Profile: ${d.risk_appetite.toUpperCase()}\nTime Horizon: ${d.time_horizon}\n\n═══════════════════════════════════════════\nSIMULATION 1 β€” Growth vs Sentiment Tradeoff\n═══════════════════════════════════════════\nSignal: ${s1.output.signal}\nConfidence: ${Math.round(s1.output.confidence * 100)}%\nConflict: ${s1.conflict_detected ? '⚠️ YES β€” fundamentals and sentiment diverge' : 'No conflict'}\nBubble Risk Score: ${s1.bubble_risk_score}/1.0\nShort-term Return: ${s1.output.short_term}\nLong-term Return: ${s1.output.long_term}\nInsight: ${s1.key_insight}\n\n═══════════════════════════════════════════\nSIMULATION 2 β€” Risk Appetite (${s2.profile.name} Investor)\n═══════════════════════════════════════════\nRecommendation: ${s2.recommendation}\nCriteria Passed: ${s2.pass_count}\nWeighted Pass Rate: ${Math.round(s2.weighted_pass_rate * 100)}%\nMax Position Size: ${s2.profile.max_position_pct}% of portfolio\nStop-Loss: ${s2.profile.stop_loss_pct}%\nTake-Profit Target: ${s2.profile.take_profit_pct}%\nInsight: ${s2.key_insight}\n\nCross-Profile Summary:\n${s2.cross_profile_comparison.map(p => ` ${p.investor_type}: ${p.recommendation} (${p.fit_score_pct}% fit)`).join('\\n')}\n\n═══════════════════════════════════════════\nSIMULATION 3 β€” Threshold Break Analysis\n═══════════════════════════════════════════\nCurrent Signal: ${s3.current_state.recommendation}\nStability Score: ${s3.stability_score}/1.0\nPrimary Driver: ${s3.sensitivity.primary_driver}\nSentiment Breaks Near: ${s3.sensitivity.sentiment_breaks_nearby}\nStress Resiliency: ${s3.sensitivity.stress_resiliency}\nStress Breach Count: ${s3.sensitivity.stress_breach_count} scenarios triggered a change\nInsight: ${s3.key_insight}\n\nStress Test Results:\n${s3.stress_tests.map(t => ` ${t.scenario}: ${t.result} (${t.changed ? '⚠️ CHANGED β€” ' + t.delta : 'βœ… unchanged'})`).join('\\n')}\n\n═══════════════════════════════════════════\nAGGREGATE VERDICT\n═══════════════════════════════════════════\nFINAL RECOMMENDATION: ${agg.final_recommendation}\nScore: ${agg.score_label}\nConfidence: ${agg.confidence_pct}%\nRisk Flags (${agg.risk_flag_count}): ${agg.risk_flags.map(f => f.flag).join(' | ') || 'None'}\n\nScenario Probabilities:\n Best Case (${agg.scenarios.best_case.probability}%): ${agg.scenarios.best_case.return_estimate}\n Expected (${agg.scenarios.expected_case.probability}%): ${agg.scenarios.expected_case.return_estimate}\n Worst Case (${agg.scenarios.worst_case.probability}%): ${agg.scenarios.worst_case.return_estimate}\n\n═══════════════════════════════════════════\nYOUR TASK\n═══════════════════════════════════════════\nReturn ONLY the following JSON β€” no markdown, no explanation outside the JSON:\n\n{\n \"executive_summary\": \"2–3 sentences covering the most important aspects of this investment case\",\n \"bull_case\": \"The single strongest argument FOR investing in this company right now\",\n \"bear_case\": \"The single most important risk that could invalidate the bullish thesis\",\n \"key_risk_narrative\": \"Explain the #1 risk in 1–2 sentences, with the specific trigger to watch for\",\n \"opportunity_narrative\": \"Explain the primary upside catalyst in 1–2 sentences\",\n \"action_plan\": {\n \"immediate\": \"Exactly what a ${d.risk_appetite} investor should do TODAY\",\n \"monitor_weekly\": \"The single most important metric to track each week\",\n \"exit_trigger\": \"The precise condition (with threshold) that would change the recommendation\",\n \"position_sizing_note\": \"Brief note on how large this position should be and why\"\n },\n \"mycroft_verdict\": \"One powerful, memorable sentence that captures the entire investment thesis for ${d.company}\",\n \"confidence_rationale\": \"Why the confidence is ${agg.confidence_pct}% β€” what is certain and what is uncertain\"\n}\nIMPORTANT: Your action_plan.immediate must strictly match \"${agg.final_recommendation}\". If HOLD β€” say hold, do not say enter.\n`;\n\nreturn [{ json: { ...d, ai_prompt: prompt } }];" }, "id": "da8a521f-bdf2-45a6-9cd3-7853da1f3f65", "name": "πŸ“ Build AI Prompt", "type": "n8n-nodes-base.code", "typeVersion": 2, "position": [ -912, 96 ] }, { "id": "a7344b8c-ff90-5b6a-b9d3-bd0f9ef9f7f2", "name": "Publication Β· Overview and setup", "type": "n8n-nodes-base.stickyNote", "typeVersion": 1, "position": [ -3152, 0 ], "parameters": { "content": "# What-If Simulation Agent\n\n## Purpose and structure\nThis folder contains the **What-If Simulation Agent** workflow. Its executable nodes use `code`, `httpRequest`, `postgres`, `respondToWebhook`, `webhook`. The table and connection map below are derived from the JSON, not an inferred guarantee of behavior.\n\n## Setup\nImport inactive. Configure your own credentials and replace REPLACE_WITH_* values. Review external destinations and Code nodes before running.\n\n## Entry points\n🎯 Webhook β€” Receive Simulation Request\n\n## Verification\nSanitized source candidate; not runtime-tested. This overview is generated from node metadata. Read the folder README for the connection map, credential types, and license.\n\n## Attribution\nNik Bear Brown; rights confirmed 2026-10-19\nMIT; preserve source notices.", "width": 700, "height": 800, "color": 1 } } ], "connections": { "🎯 Webhook β€” Receive Simulation Request": { "main": [ [ { "node": "πŸ” Parse & Validate Input", "type": "main", "index": 0 } ] ] }, "πŸ” Parse & Validate Input": { "main": [ [ { "node": "πŸ“ˆ Sim 1 β€” Growth vs Sentiment Tradeoff", "type": "main", "index": 0 } ] ] }, "πŸ“ˆ Sim 1 β€” Growth vs Sentiment Tradeoff": { "main": [ [ { "node": "βš–οΈ Sim 2 β€” Risk Appetite Simulator", "type": "main", "index": 0 } ] ] }, "βš–οΈ Sim 2 β€” Risk Appetite Simulator": { "main": [ [ { "node": "🎯 Sim 3 β€” Threshold Break Simulator", "type": "main", "index": 0 } ] ] }, "🎯 Sim 3 β€” Threshold Break Simulator": { "main": [ [ { "node": "🧠 Aggregate & Score All Simulations", "type": "main", "index": 0 } ] ] }, "🧠 Aggregate & Score All Simulations": { "main": [ [ { "node": "πŸ“ Build AI Prompt", "type": "main", "index": 0 } ] ] }, "πŸ€– Groq β€” Generate Investment Insights": { "main": [ [ { "node": "πŸ”„ Parse Groq Response", "type": "main", "index": 0 } ] ] }, "πŸ”„ Parse Groq Response": { "main": [ [ { "node": "πŸ—„οΈ Prepare DB Record", "type": "main", "index": 0 } ] ] }, "πŸ—„οΈ Prepare DB Record": { "main": [ [ { "node": "πŸ’Ύ Save to PostgreSQL", "type": "main", "index": 0 } ] ] }, "πŸ’Ύ Save to PostgreSQL": { "main": [ [ { "node": "πŸ“Š Build Final Response", "type": "main", "index": 0 } ] ] }, "πŸ“Š Build Final Response": { "main": [ [ { "node": "βœ… Respond to Webhook", "type": "main", "index": 0 } ] ] }, "πŸ“ Build AI Prompt": { "main": [ [ { "node": "πŸ€– Groq β€” Generate Investment Insights", "type": "main", "index": 0 } ] ] } }, "active": false, "settings": { "executionOrder": "v1" } }