tribunal-kit 9.2.6 → 9.2.7

package/.agent/scripts/architecture_intelligence.js

@@ -0,0 +1,3100 @@
+#!/usr/bin/env node
+/**
+ * architecture_intelligence.js — Tribunal Kit Phase 7: Architecture & Trade-Off Intelligence
+ * ===========================================================================================
+ * Core Invariant:
+ *   AN ARCHITECTURE MUST BE JUSTIFIED BY REQUIREMENTS, CONSTRAINTS, AND EVIDENCE.
+ *   NO ARCHITECTURAL CHOICE SHOULD BE MADE SOLELY BECAUSE IT IS FAMILIAR, POPULAR,
+ *   MODERN, OR EASY FOR THE MODEL TO GENERATE.
+ *
+ * Pipeline:
+ *   TASK → REQUIREMENTS → CONSTRAINTS → CONSIDERATIONS → ARCHITECTURE DISCOVERY
+ *   → CANDIDATE GENERATION → VALIDATION → TRADE-OFF ANALYSIS → FAILURE ANALYSIS
+ *   → COST/PERF/SECURITY/OPS ANALYSIS → DECISION RECORD → VERIFICATION
+ *
+ * Implements Phase 7 Sections 4–46.
+ */
+
+'use strict';
+
+const crypto = require('crypto');
+
+// ============================================================================
+// 1. REQUIREMENT INTELLIGENCE (SECTIONS 4–5)
+// ============================================================================
+
+const REQUIREMENT_TYPES = {
+  FUNCTIONAL: 'functional',
+  NON_FUNCTIONAL: 'non_functional',
+  CONSTRAINT: 'constraint',
+  PREFERENCE: 'preference',
+  ASSUMPTION: 'assumption',
+  UNKNOWN: 'unknown',
+};
+
+const CONFIDENCE_LEVELS = {
+  EXPLICIT: 'explicit',
+  INFERRED: 'inferred',
+  ASSUMED: 'assumed',
+  UNKNOWN: 'unknown',
+};
+
+/**
+ * Functional requirement detection patterns.
+ * Each pattern maps a keyword domain to a functional requirement.
+ */
+const FUNCTIONAL_PATTERNS = [
+  {
+    match: /(?:messag|chat|inbox|conversation|dm|direct message)/i,
+    req: 'messaging',
+    domain: 'communication',
+  },
+  {
+    match: /(?:presence|online|offline|status|active user)/i,
+    req: 'presence',
+    domain: 'communication',
+  },
+  {
+    match: /(?:history|archive|log|conversation record)/i,
+    req: 'conversation history',
+    domain: 'data',
+  },
+  {
+    match: /(?:payment|checkout|charge|billing|invoice|subscription)/i,
+    req: 'payment processing',
+    domain: 'financial',
+  },
+  {
+    match: /(?:auth|login|signup|register|sso|oauth|credential)/i,
+    req: 'authentication',
+    domain: 'security',
+  },
+  {
+    match: /(?:permission|role|rbac|acl|access control|authorization)/i,
+    req: 'authorization',
+    domain: 'security',
+  },
+  {
+    match: /(?:upload|file|attachment|media|image|video)/i,
+    req: 'file management',
+    domain: 'storage',
+  },
+  { match: /(?:search|filter|query|lookup|find|full.text)/i, req: 'search', domain: 'data' },
+  {
+    match: /(?:notification|alert|push|email|sms|webhook)/i,
+    req: 'notifications',
+    domain: 'communication',
+  },
+  {
+    match: /(?:dashboard|report|analytic|metric|chart|graph)/i,
+    req: 'analytics & reporting',
+    domain: 'data',
+  },
+  { match: /(?:feed|timeline|stream|wall|activity)/i, req: 'activity feed', domain: 'social' },
+  { match: /(?:recommend|suggest|personalize|ml|ai model)/i, req: 'recommendations', domain: 'ai' },
+  {
+    match: /(?:collaboration|realtime|collab|edit.*together|shared.*doc)/i,
+    req: 'real-time collaboration',
+    domain: 'collaboration',
+  },
+  {
+    match: /(?:crud|create|read|update|delete|manage.*record)/i,
+    req: 'CRUD operations',
+    domain: 'data',
+  },
+  { match: /(?:import|export|csv|excel|bulk)/i, req: 'data import/export', domain: 'data' },
+  {
+    match: /(?:schedule|cron|timer|queue|background|worker|job)/i,
+    req: 'background processing',
+    domain: 'compute',
+  },
+  {
+    match: /(?:tenant|multi.?tenant|workspace|organization)/i,
+    req: 'multi-tenancy',
+    domain: 'isolation',
+  },
+  {
+    match: /(?:agent|tool|llm|prompt|chain|rag|vector)/i,
+    req: 'AI agent integration',
+    domain: 'ai',
+  },
+  { match: /(?:api|endpoint|rest|graphql|grpc|gateway)/i, req: 'API layer', domain: 'integration' },
+  { match: /(?:cache|redis|memcache|cdn)/i, req: 'caching layer', domain: 'performance' },
+];
+
+/**
+ * Non-functional requirement detection patterns.
+ */
+const NON_FUNCTIONAL_PATTERNS = [
+  {
+    match: /(?:low.?latency|fast|responsive|sub.?\d+ms|p99|p95|latency)/i,
+    req: 'low latency',
+    category: 'performance',
+  },
+  {
+    match: /(?:high.?availability|uptime|99\.9|five.?nine|four.?nine|ha\b)/i,
+    req: 'high availability',
+    category: 'reliability',
+  },
+  {
+    match: /(?:scal|horizontal|vertical|elastic|auto.?scale)/i,
+    req: 'scalability',
+    category: 'scalability',
+  },
+  {
+    match: /(?:secure|encrypt|tls|ssl|zero.?trust|penetration)/i,
+    req: 'security',
+    category: 'security',
+  },
+  {
+    match: /(?:reliable|fault.?tolerant|resilient|durable|crash.?safe)/i,
+    req: 'reliability',
+    category: 'reliability',
+  },
+  {
+    match: /(?:consistent|acid|lineariz|serial|strong.?consisten)/i,
+    req: 'strong consistency',
+    category: 'consistency',
+  },
+  {
+    match: /(?:eventual|async|lazy|background.?sync)/i,
+    req: 'eventual consistency',
+    category: 'consistency',
+  },
+  {
+    match: /(?:observab|monitor|trace|log|metric|opentelemetry)/i,
+    req: 'observability',
+    category: 'operations',
+  },
+  {
+    match: /(?:test|ci|cd|pipeline|deploy|rollback)/i,
+    req: 'CI/CD & deployability',
+    category: 'operations',
+  },
+  {
+    match: /(?:maintain|clean|modular|readable|document)/i,
+    req: 'maintainability',
+    category: 'quality',
+  },
+  {
+    match: /(?:audit|compliance|gdpr|hipaa|sox|pci|regulatory)/i,
+    req: 'compliance & auditability',
+    category: 'governance',
+  },
+  {
+    match: /(?:cost|budget|cheap|expensive|billing.?optim)/i,
+    req: 'cost efficiency',
+    category: 'cost',
+  },
+];
+
+/**
+ * Constraint extraction patterns.
+ */
+const CONSTRAINT_PATTERNS = [
+  {
+    match: /(\d[\d,.]*)\s*(?:user|customer|account|subscriber)/i,
+    req: 'user count',
+    extractor: m => m[1],
+    category: 'scale',
+  },
+  {
+    match: /(\d[\d,.]*)\s*(?:req|request|rps|qps|tps)(?:\/s|\/sec|\/second)?/i,
+    req: 'request throughput',
+    extractor: m => m[1],
+    category: 'performance',
+  },
+  {
+    match: /(\d[\d,.]*)\s*(?:gb|tb|pb|mb)\s*(?:data|storage|size)/i,
+    req: 'data volume',
+    extractor: m => m[1],
+    category: 'storage',
+  },
+  {
+    match: /(?:aws|gcp|azure|vercel|cloudflare|heroku|digital.?ocean|fly\.io)/i,
+    req: 'cloud provider',
+    category: 'infrastructure',
+  },
+  {
+    match: /(?:postgres|mysql|mongo|dynamo|redis|sqlite|cockroach|supabase|planetscale|neon)/i,
+    req: 'database technology',
+    category: 'technology',
+  },
+  {
+    match: /(?:react|vue|angular|svelte|next|nuxt|remix|astro)/i,
+    req: 'frontend framework',
+    category: 'technology',
+  },
+  {
+    match: /(?:node|python|go|rust|java|c#|\.net|ruby|elixir|php)/i,
+    req: 'backend language',
+    category: 'technology',
+  },
+  {
+    match: /(?:docker|kubernetes|k8s|serverless|lambda|container)/i,
+    req: 'deployment platform',
+    category: 'infrastructure',
+  },
+  {
+    match: /(?:team of (\d+)|(\d+)\s*dev|solo|one.?person|small team|large team)/i,
+    req: 'team size',
+    category: 'team',
+  },
+  {
+    match: /(?:deadline|timeline|sprint|week|month|quarter|mvp|prototype|poc)/i,
+    req: 'timeline',
+    category: 'schedule',
+  },
+  {
+    match: /(?:gdpr|hipaa|sox|pci|ccpa|iso.?27001|fedramp)/i,
+    req: 'compliance requirement',
+    category: 'governance',
+  },
+  {
+    match: /(?:data.?residen|sovereign|eu.?only|us.?only|region.?restrict)/i,
+    req: 'data residency',
+    category: 'governance',
+  },
+  {
+    match: /(?:existing|legacy|migrate|current system|brownfield)/i,
+    req: 'existing system',
+    category: 'migration',
+  },
+  {
+    match: /(?:budget|cost.?limit|\$\d+|free.?tier|startup)/i,
+    req: 'budget constraint',
+    category: 'cost',
+  },
+  {
+    match: /(?:real.?time|sub.?second|instant|live|streaming)/i,
+    req: 'real-time requirement',
+    category: 'performance',
+  },
+  {
+    match: /(?:offline|pwa|local.?first|sync)/i,
+    req: 'offline capability',
+    category: 'availability',
+  },
+];
+
+/**
+ * Extract structured requirements from a natural-language task description.
+ *
+ * @param {string} taskQuery
+ * @returns {Object} Structured requirements
+ */
+function extractRequirements(taskQuery) {
+  const q = (taskQuery || '').trim();
+  if (!q)
+    return {
+      functional: [],
+      non_functional: [],
+      constraints: [],
+      preferences: [],
+      assumptions: [],
+      unknowns: [],
+      raw_task: q,
+    };
+
+  const functional = [];
+  const nonFunctional = [];
+  const constraints = [];
+  const preferences = [];
+  const assumptions = [];
+  const unknowns = [];
+  const seen = new Set();
+
+  // Extract functional requirements
+  for (const p of FUNCTIONAL_PATTERNS) {
+    if (p.match.test(q) && !seen.has(p.req)) {
+      seen.add(p.req);
+      functional.push({
+        requirement: p.req,
+        source: 'task_description',
+        type: REQUIREMENT_TYPES.FUNCTIONAL,
+        confidence: CONFIDENCE_LEVELS.EXPLICIT,
+        domain: p.domain,
+        evidence: [{ type: 'keyword_match', detail: `Detected "${p.req}" from task description` }],
+      });
+    }
+  }
+
+  // Extract non-functional requirements
+  for (const p of NON_FUNCTIONAL_PATTERNS) {
+    if (p.match.test(q) && !seen.has(p.req)) {
+      seen.add(p.req);
+      nonFunctional.push({
+        requirement: p.req,
+        source: 'task_description',
+        type: REQUIREMENT_TYPES.NON_FUNCTIONAL,
+        confidence: CONFIDENCE_LEVELS.EXPLICIT,
+        category: p.category,
+        evidence: [{ type: 'keyword_match', detail: `Detected "${p.req}" from task description` }],
+      });
+    }
+  }
+
+  // Extract constraints
+  for (const p of CONSTRAINT_PATTERNS) {
+    const m = q.match(p.match);
+    if (m && !seen.has(p.req)) {
+      seen.add(p.req);
+      constraints.push({
+        requirement: p.req,
+        source: 'task_description',
+        type: REQUIREMENT_TYPES.CONSTRAINT,
+        confidence: CONFIDENCE_LEVELS.EXPLICIT,
+        category: p.category,
+        value: p.extractor ? p.extractor(m) : m[0],
+        evidence: [{ type: 'keyword_match', detail: `Detected "${p.req}" constraint: ${m[0]}` }],
+      });
+    }
+  }
+
+  // Infer implicit requirements based on functional requirements
+  if (functional.some(f => f.domain === 'financial') && !seen.has('idempotency')) {
+    seen.add('idempotency');
+    nonFunctional.push({
+      requirement: 'idempotency',
+      source: 'inferred_from_financial_domain',
+      type: REQUIREMENT_TYPES.NON_FUNCTIONAL,
+      confidence: CONFIDENCE_LEVELS.INFERRED,
+      category: 'reliability',
+      evidence: [
+        {
+          type: 'domain_inference',
+          detail: 'Financial transactions require idempotency to prevent duplicate charges',
+        },
+      ],
+    });
+  }
+
+  if (functional.some(f => f.domain === 'financial') && !seen.has('auditability')) {
+    seen.add('auditability');
+    nonFunctional.push({
+      requirement: 'auditability',
+      source: 'inferred_from_financial_domain',
+      type: REQUIREMENT_TYPES.NON_FUNCTIONAL,
+      confidence: CONFIDENCE_LEVELS.INFERRED,
+      category: 'governance',
+      evidence: [
+        {
+          type: 'domain_inference',
+          detail: 'Financial systems require audit trails for regulatory compliance',
+        },
+      ],
+    });
+  }
+
+  if (functional.some(f => f.domain === 'communication') && !seen.has('low latency')) {
+    seen.add('low latency');
+    nonFunctional.push({
+      requirement: 'low latency',
+      source: 'inferred_from_communication_domain',
+      type: REQUIREMENT_TYPES.NON_FUNCTIONAL,
+      confidence: CONFIDENCE_LEVELS.INFERRED,
+      category: 'performance',
+      evidence: [
+        {
+          type: 'domain_inference',
+          detail: 'Communication systems require low latency for user experience',
+        },
+      ],
+    });
+  }
+
+  // Identify unknowns based on what's NOT specified
+  const unknownChecks = [
+    { condition: !constraints.some(c => c.category === 'scale'), unknown: 'peak concurrent users' },
+    {
+      condition: !constraints.some(c => c.req === 'request throughput'),
+      unknown: 'request rate / throughput',
+    },
+    {
+      condition: !constraints.some(c => c.category === 'storage'),
+      unknown: 'data retention period',
+    },
+    {
+      condition: !constraints.some(c => c.req === 'data residency'),
+      unknown: 'geographic distribution',
+    },
+    {
+      condition: !constraints.some(c => c.category === 'team'),
+      unknown: 'team size and expertise',
+    },
+    { condition: !constraints.some(c => c.category === 'cost'), unknown: 'budget' },
+    { condition: !constraints.some(c => c.category === 'schedule'), unknown: 'timeline' },
+  ];
+
+  for (const check of unknownChecks) {
+    if (check.condition) {
+      unknowns.push({
+        requirement: check.unknown,
+        type: REQUIREMENT_TYPES.UNKNOWN,
+        confidence: CONFIDENCE_LEVELS.UNKNOWN,
+        evidence: [],
+      });
+    }
+  }
+
+  return {
+    functional,
+    non_functional: nonFunctional,
+    constraints,
+    preferences,
+    assumptions,
+    unknowns,
+    raw_task: q,
+    total: functional.length + nonFunctional.length + constraints.length,
+  };
+}
+
+// ============================================================================
+// 2. CONSTRAINT EXTRACTION (SECTION 6)
+// ============================================================================
+
+/**
+ * Extract and categorize all constraints from task + requirements.
+ *
+ * @param {string} taskQuery
+ * @param {Object} requirements
+ * @returns {Object} Structured constraints
+ */
+function extractConstraints(taskQuery, requirements) {
+  const q = (taskQuery || '').toLowerCase();
+  const existingConstraints = (requirements && requirements.constraints) || [];
+
+  const allConstraints = [...existingConstraints];
+  const categories = new Set(existingConstraints.map(c => c.category));
+
+  // Budget constraint inference
+  if (
+    !categories.has('cost') &&
+    /(?:startup|bootstrap|mvp|poc|prototype|free.?tier|hobby)/i.test(q)
+  ) {
+    allConstraints.push({
+      requirement: 'limited budget',
+      type: REQUIREMENT_TYPES.CONSTRAINT,
+      confidence: CONFIDENCE_LEVELS.INFERRED,
+      category: 'cost',
+      value: 'low',
+      evidence: [
+        {
+          type: 'context_inference',
+          detail: 'Startup/MVP/prototype context suggests limited budget',
+        },
+      ],
+    });
+  }
+
+  // Team expertise inference
+  if (
+    !categories.has('team') &&
+    /(?:solo|one.?person|small team|junior|learning|first time)/i.test(q)
+  ) {
+    allConstraints.push({
+      requirement: 'limited team capacity',
+      type: REQUIREMENT_TYPES.CONSTRAINT,
+      confidence: CONFIDENCE_LEVELS.INFERRED,
+      category: 'team',
+      value: 'small',
+      evidence: [
+        {
+          type: 'context_inference',
+          detail: 'Small team or solo developer context constrains operational complexity',
+        },
+      ],
+    });
+  }
+
+  // Operational maturity inference
+  if (
+    !categories.has('operations') &&
+    /(?:no.?devops|no.?infra|managed|serverless|simple.?deploy)/i.test(q)
+  ) {
+    allConstraints.push({
+      requirement: 'limited operational maturity',
+      type: REQUIREMENT_TYPES.CONSTRAINT,
+      confidence: CONFIDENCE_LEVELS.INFERRED,
+      category: 'operations',
+      value: 'low',
+      evidence: [
+        {
+          type: 'context_inference',
+          detail: 'No DevOps expertise mentioned; prefer simpler operational models',
+        },
+      ],
+    });
+  }
+
+  return {
+    constraints: allConstraints,
+    categories: [...new Set(allConstraints.map(c => c.category))],
+    total: allConstraints.length,
+  };
+}
+
+// ============================================================================
+// 3. ARCHITECTURE PATTERN REGISTRY (SECTION 8)
+// ============================================================================
+
+const ARCHITECTURE_PATTERNS = {
+  monolith: {
+    pattern: 'Monolith',
+    use_cases: ['small-to-medium applications', 'early-stage products', 'simple domain'],
+    strengths: [
+      'simple deployment',
+      'simple debugging',
+      'low operational cost',
+      'strong consistency',
+      'fast development',
+      'easy testing',
+    ],
+    weaknesses: [
+      'scaling limitations',
+      'deployment coupling',
+      'technology lock-in',
+      'team coupling',
+    ],
+    failure_modes: ['single point of failure', 'memory exhaustion', 'deployment blocks'],
+    operational_cost: 'low',
+    scaling_characteristics: {
+      horizontal: 'limited',
+      vertical: 'primary',
+      data_partitioning: 'difficult',
+    },
+    security_considerations: ['single trust boundary', 'shared memory space'],
+    related_patterns: ['modular_monolith', 'layered'],
+    anti_patterns: ['god_service'],
+    required_capabilities: ['single deployment pipeline'],
+  },
+  modular_monolith: {
+    pattern: 'Modular Monolith',
+    use_cases: [
+      'medium applications',
+      'clear domain boundaries',
+      'team growth path',
+      'payment systems',
+      'SaaS platforms',
+    ],
+    strengths: [
+      'domain isolation within single deployment',
+      'simple operations',
+      'strong consistency',
+      'easy refactoring to services later',
+      'independent module development',
+    ],
+    weaknesses: [
+      'requires discipline to maintain module boundaries',
+      'shared database',
+      'single deployment unit',
+    ],
+    failure_modes: ['boundary erosion', 'shared state leakage', 'deployment coupling persists'],
+    operational_cost: 'low',
+    scaling_characteristics: {
+      horizontal: 'moderate',
+      vertical: 'primary',
+      data_partitioning: 'module-scoped',
+    },
+    security_considerations: ['module-level access control', 'shared process space'],
+    related_patterns: ['monolith', 'hexagonal', 'microservices'],
+    anti_patterns: ['distributed_monolith'],
+    required_capabilities: ['module boundary enforcement', 'internal API contracts'],
+  },
+  microservices: {
+    pattern: 'Microservices',
+    use_cases: [
+      'large-scale systems',
+      'multiple teams',
+      'independent deployment',
+      'polyglot technology',
+    ],
+    strengths: [
+      'independent deployment',
+      'technology flexibility',
+      'team autonomy',
+      'independent scaling',
+      'fault isolation',
+    ],
+    weaknesses: [
+      'distributed system complexity',
+      'network latency',
+      'data consistency challenges',
+      'operational overhead',
+      'debugging difficulty',
+    ],
+    failure_modes: [
+      'cascading failures',
+      'network partitions',
+      'distributed deadlocks',
+      'service discovery failures',
+      'data inconsistency',
+    ],
+    operational_cost: 'high',
+    scaling_characteristics: {
+      horizontal: 'excellent',
+      vertical: 'per-service',
+      data_partitioning: 'per-service',
+    },
+    security_considerations: [
+      'service-to-service auth',
+      'network segmentation',
+      'secret distribution',
+      'multiple trust boundaries',
+    ],
+    related_patterns: ['soa', 'api_gateway', 'service_mesh', 'saga'],
+    anti_patterns: ['premature_microservices', 'distributed_monolith', 'chatty_services'],
+    required_capabilities: [
+      'service discovery',
+      'distributed tracing',
+      'circuit breakers',
+      'container orchestration',
+      'CI/CD per service',
+    ],
+  },
+  serverless: {
+    pattern: 'Serverless',
+    use_cases: [
+      'event-driven workloads',
+      'variable traffic',
+      'rapid prototyping',
+      'webhook handlers',
+      'scheduled jobs',
+    ],
+    strengths: [
+      'zero server management',
+      'auto-scaling',
+      'pay-per-use',
+      'fast deployment',
+      'no idle cost',
+    ],
+    weaknesses: [
+      'cold starts',
+      'execution time limits',
+      'vendor lock-in',
+      'debugging difficulty',
+      'limited local testing',
+      'stateless constraint',
+    ],
+    failure_modes: [
+      'cold start latency',
+      'timeout errors',
+      'concurrent execution limits',
+      'vendor outages',
+    ],
+    operational_cost: 'very_low_to_variable',
+    scaling_characteristics: {
+      horizontal: 'automatic',
+      vertical: 'limited_by_runtime',
+      data_partitioning: 'external',
+    },
+    security_considerations: [
+      'IAM roles',
+      'environment variable secrets',
+      'function-level permissions',
+    ],
+    related_patterns: ['event_driven', 'api_gateway'],
+    anti_patterns: ['over_engineering'],
+    required_capabilities: ['managed database', 'event source integration'],
+  },
+  event_driven: {
+    pattern: 'Event-Driven',
+    use_cases: [
+      'async workflows',
+      'decoupled systems',
+      'audit trails',
+      'real-time processing',
+      'CQRS',
+    ],
+    strengths: [
+      'loose coupling',
+      'async processing',
+      'audit trail',
+      'replay capability',
+      'temporal decoupling',
+    ],
+    weaknesses: [
+      'eventual consistency',
+      'debugging complexity',
+      'event ordering',
+      'schema evolution',
+      'idempotency requirement',
+    ],
+    failure_modes: [
+      'event loss',
+      'duplicate events',
+      'out-of-order processing',
+      'poison pills',
+      'consumer lag',
+    ],
+    operational_cost: 'medium_to_high',
+    scaling_characteristics: {
+      horizontal: 'excellent',
+      vertical: 'per-consumer',
+      data_partitioning: 'partition-based',
+    },
+    security_considerations: [
+      'event payload encryption',
+      'consumer authorization',
+      'event source verification',
+    ],
+    related_patterns: ['cqrs', 'event_sourcing', 'saga', 'transactional_outbox'],
+    anti_patterns: ['dual_write'],
+    required_capabilities: [
+      'message broker',
+      'dead-letter queue',
+      'idempotent consumers',
+      'schema registry',
+    ],
+  },
+  cqrs: {
+    pattern: 'CQRS',
+    use_cases: [
+      'read-heavy systems',
+      'complex queries',
+      'separate read/write scaling',
+      'event sourcing complement',
+    ],
+    strengths: ['optimized read/write models', 'independent scaling', 'query flexibility'],
+    weaknesses: [
+      'increased complexity',
+      'eventual consistency between read/write',
+      'synchronization logic',
+    ],
+    failure_modes: ['read model staleness', 'sync failures', 'write model corruption'],
+    operational_cost: 'medium',
+    scaling_characteristics: {
+      horizontal: 'per-model',
+      vertical: 'per-model',
+      data_partitioning: 'read/write split',
+    },
+    security_considerations: ['separate read/write authorization'],
+    related_patterns: ['event_driven', 'event_sourcing'],
+    anti_patterns: ['over_engineering'],
+    required_capabilities: ['separate data stores or views', 'sync mechanism'],
+  },
+  event_sourcing: {
+    pattern: 'Event Sourcing',
+    use_cases: [
+      'audit-critical systems',
+      'financial ledgers',
+      'temporal queries',
+      'regulatory compliance',
+    ],
+    strengths: [
+      'complete audit trail',
+      'temporal queries',
+      'replay capability',
+      'debugging via event replay',
+    ],
+    weaknesses: ['storage growth', 'complexity', 'snapshot management', 'schema evolution'],
+    failure_modes: ['event store corruption', 'projection failures', 'snapshot inconsistency'],
+    operational_cost: 'high',
+    scaling_characteristics: {
+      horizontal: 'partition-based',
+      vertical: 'moderate',
+      data_partitioning: 'aggregate-based',
+    },
+    security_considerations: ['immutable event store access control', 'PII in events'],
+    related_patterns: ['cqrs', 'event_driven'],
+    anti_patterns: [],
+    required_capabilities: ['event store', 'projection engine', 'snapshot mechanism'],
+  },
+  hexagonal: {
+    pattern: 'Hexagonal (Ports & Adapters)',
+    use_cases: [
+      'domain-driven applications',
+      'testable business logic',
+      'infrastructure-agnostic core',
+    ],
+    strengths: ['infrastructure independence', 'highly testable', 'clear domain boundaries'],
+    weaknesses: ['more boilerplate', 'learning curve', 'over-abstraction risk'],
+    failure_modes: ['port/adapter mismatch', 'leaky abstractions'],
+    operational_cost: 'low',
+    scaling_characteristics: {
+      horizontal: 'inherited',
+      vertical: 'inherited',
+      data_partitioning: 'inherited',
+    },
+    security_considerations: ['adapter-level validation'],
+    related_patterns: ['clean_architecture', 'modular_monolith'],
+    anti_patterns: ['over_engineering'],
+    required_capabilities: ['dependency injection'],
+  },
+  api_gateway: {
+    pattern: 'API Gateway',
+    use_cases: [
+      'microservices entry point',
+      'rate limiting',
+      'auth consolidation',
+      'request routing',
+    ],
+    strengths: [
+      'centralized cross-cutting concerns',
+      'client simplification',
+      'protocol translation',
+    ],
+    weaknesses: ['single point of failure', 'added latency', 'deployment bottleneck'],
+    failure_modes: ['gateway overload', 'routing errors', 'configuration drift'],
+    operational_cost: 'medium',
+    scaling_characteristics: {
+      horizontal: 'excellent',
+      vertical: 'moderate',
+      data_partitioning: 'not_applicable',
+    },
+    security_considerations: ['auth termination point', 'rate limiting', 'DDoS protection'],
+    related_patterns: ['microservices', 'bff'],
+    anti_patterns: ['god_service'],
+    required_capabilities: ['load balancer', 'health checks'],
+  },
+  saga: {
+    pattern: 'Saga (Distributed Transactions)',
+    use_cases: [
+      'cross-service transactions',
+      'long-running workflows',
+      'compensating transactions',
+    ],
+    strengths: ['distributed consistency without 2PC', 'fault tolerance', 'eventual consistency'],
+    weaknesses: ['compensation complexity', 'debugging difficulty', 'partial failure handling'],
+    failure_modes: ['compensation failures', 'orphaned sagas', 'timeout deadlocks'],
+    operational_cost: 'medium_to_high',
+    scaling_characteristics: {
+      horizontal: 'per-step',
+      vertical: 'per-step',
+      data_partitioning: 'per-service',
+    },
+    security_considerations: ['step authorization', 'compensation authorization'],
+    related_patterns: ['event_driven', 'transactional_outbox', 'microservices'],
+    anti_patterns: ['distributed_transaction_without_need'],
+    required_capabilities: [
+      'saga orchestrator or choreography',
+      'compensating actions',
+      'idempotent steps',
+    ],
+  },
+  transactional_outbox: {
+    pattern: 'Transactional Outbox',
+    use_cases: ['reliable event publishing', 'dual write prevention', 'at-least-once delivery'],
+    strengths: ['atomicity between DB write and event publish', 'reliable delivery'],
+    weaknesses: [
+      'polling overhead',
+      'outbox table growth',
+      'at-least-once semantics require idempotent consumers',
+    ],
+    failure_modes: ['outbox reader lag', 'duplicate delivery'],
+    operational_cost: 'low_to_medium',
+    scaling_characteristics: {
+      horizontal: 'moderate',
+      vertical: 'moderate',
+      data_partitioning: 'per-service',
+    },
+    security_considerations: ['outbox table access control'],
+    related_patterns: ['event_driven', 'saga'],
+    anti_patterns: ['dual_write'],
+    required_capabilities: ['database with transactional support', 'outbox reader/poller'],
+  },
+  streaming: {
+    pattern: 'Streaming Architecture',
+    use_cases: [
+      'real-time analytics',
+      'IoT data processing',
+      'event processing',
+      'log aggregation',
+    ],
+    strengths: ['real-time processing', 'high throughput', 'temporal windowing'],
+    weaknesses: [
+      'complexity',
+      'ordering challenges',
+      'exactly-once difficulty',
+      'operational cost',
+    ],
+    failure_modes: ['backpressure', 'out-of-order events', 'late data'],
+    operational_cost: 'high',
+    scaling_characteristics: {
+      horizontal: 'partition-based',
+      vertical: 'per-consumer',
+      data_partitioning: 'topic/partition',
+    },
+    security_considerations: ['stream encryption', 'consumer group authorization'],
+    related_patterns: ['event_driven', 'lambda_architecture', 'kappa_architecture'],
+    anti_patterns: ['over_engineering'],
+    required_capabilities: ['stream processing framework', 'message broker with partitioning'],
+  },
+  bff: {
+    pattern: 'Backend for Frontend',
+    use_cases: ['multiple client types', 'mobile + web + API', 'client-specific aggregation'],
+    strengths: ['client-optimized API', 'reduced over-fetching', 'client-specific caching'],
+    weaknesses: ['code duplication across BFFs', 'added services', 'sync overhead'],
+    failure_modes: ['BFF-backend desync', 'BFF proliferation'],
+    operational_cost: 'medium',
+    scaling_characteristics: {
+      horizontal: 'per-BFF',
+      vertical: 'per-BFF',
+      data_partitioning: 'not_applicable',
+    },
+    security_considerations: ['per-client auth', 'data filtering per client type'],
+    related_patterns: ['api_gateway', 'microservices'],
+    anti_patterns: ['god_service'],
+    required_capabilities: ['per-client deployment pipelines'],
+  },
+  distributed_workers: {
+    pattern: 'Distributed Workers',
+    use_cases: [
+      'background processing',
+      'task queues',
+      'parallel computation',
+      'email/notification pipelines',
+    ],
+    strengths: ['horizontal scaling', 'fault isolation', 'async processing'],
+    weaknesses: ['monitoring complexity', 'dead-letter handling', 'worker coordination'],
+    failure_modes: ['worker crash', 'poison pills', 'starvation', 'thundering herd'],
+    operational_cost: 'medium',
+    scaling_characteristics: {
+      horizontal: 'excellent',
+      vertical: 'per-worker',
+      data_partitioning: 'queue-based',
+    },
+    security_considerations: ['worker credential management', 'job payload validation'],
+    related_patterns: ['event_driven', 'serverless'],
+    anti_patterns: ['unbounded_retry'],
+    required_capabilities: ['message queue', 'dead-letter queue', 'health monitoring'],
+  },
+  batch: {
+    pattern: 'Batch Architecture',
+    use_cases: ['ETL pipelines', 'data warehousing', 'report generation', 'nightly processing'],
+    strengths: ['high throughput', 'simple error handling', 'resource scheduling'],
+    weaknesses: ['high latency', 'resource spikes', 'stale data'],
+    failure_modes: ['timeout', 'partial failure', 'data corruption on restart'],
+    operational_cost: 'low_to_medium',
+    scaling_characteristics: {
+      horizontal: 'partition-based',
+      vertical: 'moderate',
+      data_partitioning: 'time/partition-based',
+    },
+    security_considerations: ['data access scoping', 'temporary credential lifecycle'],
+    related_patterns: ['streaming', 'distributed_workers'],
+    anti_patterns: [],
+    required_capabilities: ['scheduler', 'data pipeline framework'],
+  },
+};
+
+/**
+ * Retrieve all architecture patterns or a specific one.
+ *
+ * @param {string} [patternName]
+ * @returns {Object|Object[]}
+ */
+function getArchitecturePatterns(patternName) {
+  if (patternName) {
+    const key = patternName.toLowerCase().replace(/[\s-]+/g, '_');
+    return ARCHITECTURE_PATTERNS[key] || null;
+  }
+  return { ...ARCHITECTURE_PATTERNS };
+}
+
+// ============================================================================
+// 4. ANTI-PATTERN REGISTRY (SECTION 9)
+// ============================================================================
+
+const ANTI_PATTERNS = {
+  premature_microservices: {
+    name: 'Premature Microservices',
+    description:
+      'Adopting microservices before the domain is understood or the team can operate them.',
+    detection: /(?:microservice|micro.service)/i,
+    contraindications: ['small team', 'mvp', 'prototype', 'solo', 'startup', 'simple', 'crud'],
+    severity: 'HIGH',
+    rationale:
+      'Microservices introduce distributed system complexity (network failures, data consistency, deployment coordination) that is unjustified when a modular monolith suffices.',
+    alternative: 'modular_monolith',
+  },
+  distributed_monolith: {
+    name: 'Distributed Monolith',
+    description: 'Multiple services that must be deployed together and share a database.',
+    detection:
+      /(?:distributed.monolith|shared.database.*microservice|microservice.*shared.database)/i,
+    contraindications: [],
+    severity: 'CRITICAL',
+    rationale:
+      'Combines the complexity of distributed systems with none of the benefits (independent deployment, fault isolation).',
+    alternative: 'modular_monolith',
+  },
+  shared_database_coupling: {
+    name: 'Shared Database Coupling',
+    description: 'Multiple services directly accessing the same database tables.',
+    detection: /(?:shared.database|common.db|same.database)/i,
+    contraindications: [],
+    severity: 'HIGH',
+    rationale:
+      'Creates implicit coupling: schema changes in one service break others. Prevents independent scaling and deployment.',
+    alternative: 'transactional_outbox',
+  },
+  synchronous_chain: {
+    name: 'Synchronous Chain',
+    description: 'Long chains of synchronous service-to-service calls.',
+    detection: /(?:synchronous.chain|sync.call.*chain|chain.*sync)/i,
+    contraindications: [],
+    severity: 'HIGH',
+    rationale:
+      'Latency compounds multiplicatively; any service failure cascades through the entire chain.',
+    alternative: 'event_driven',
+  },
+  cache_as_source_of_truth: {
+    name: 'Cache as Source of Truth',
+    description: 'Using cache as the primary data store instead of a durable database.',
+    detection: /(?:cache.as.source|cache.only|redis.only.store)/i,
+    contraindications: [],
+    severity: 'CRITICAL',
+    rationale:
+      'Caches are ephemeral by design; eviction, restart, or memory pressure causes data loss.',
+    alternative: 'modular_monolith',
+  },
+  dual_write: {
+    name: 'Dual Write',
+    description: 'Writing to two systems (DB + message broker) without atomicity.',
+    detection: /(?:dual.write|write.*both|db.*and.*queue.*write)/i,
+    contraindications: [],
+    severity: 'HIGH',
+    rationale:
+      'If either write fails, the systems become permanently inconsistent. Use transactional outbox instead.',
+    alternative: 'transactional_outbox',
+  },
+  unbounded_retry: {
+    name: 'Unbounded Retry',
+    description: 'Retrying failed operations without backoff, jitter, or circuit breaking.',
+    detection: /(?:retry.*forever|unlimited.retry|no.backoff)/i,
+    contraindications: [],
+    severity: 'HIGH',
+    rationale: 'Creates retry storms that overwhelm recovering services and amplify failures.',
+    alternative: null,
+  },
+  over_engineering: {
+    name: 'Over-Engineering',
+    description: 'Using complex patterns (Kafka, K8s, event sourcing) for simple problems.',
+    detection: /(?:kafka|kubernetes|k8s|event.sourc|cqrs)/i,
+    contraindications: [
+      'high throughput',
+      'million',
+      'distributed',
+      'multi-region',
+      'audit trail',
+      'regulatory',
+      'compliance',
+      'large team',
+      'real-time analytics',
+    ],
+    severity: 'MEDIUM',
+    rationale:
+      'Complexity without corresponding requirements creates operational burden, slows development, and introduces unnecessary failure modes.',
+    alternative: 'modular_monolith',
+  },
+  under_engineering: {
+    name: 'Under-Engineering',
+    description:
+      'Missing critical engineering for the requirements (no transactions for payments, no auth for sensitive data).',
+    detection: null, // Detected by absence, not by keyword
+    contraindications: [],
+    severity: 'CRITICAL',
+    rationale:
+      'Missing engineering fundamentals causes data corruption, security breaches, and financial loss.',
+    alternative: null,
+  },
+};
+
+/**
+ * Detect architecture anti-patterns in a task description.
+ *
+ * @param {string} taskQuery
+ * @param {Object} [requirements]
+ * @returns {Array<Object>} Detected anti-patterns with context
+ */
+function detectAntiPatterns(taskQuery, requirements) {
+  const q = (taskQuery || '').toLowerCase();
+  const detected = [];
+  const detectedKeys = new Set();
+
+  for (const [key, ap] of Object.entries(ANTI_PATTERNS)) {
+    if (!ap.detection) continue;
+    if (!ap.detection.test(q)) continue;
+
+    // Check contraindications — if task mentions a valid use case, this isn't an anti-pattern
+    const hasValidContext = ap.contraindications.some(ci => q.includes(ci));
+    if (hasValidContext && key === 'over_engineering') continue; // Don't flag overengineering when context justifies it
+
+    // For premature_microservices, flag only if contraindications match (e.g. "microservices" + "startup")
+    if (key === 'premature_microservices' && !hasValidContext) continue;
+
+    if (key === 'premature_microservices') {
+      detected.push({
+        anti_pattern: ap.name,
+        severity: ap.severity,
+        rationale: ap.rationale,
+        context_aware: true,
+        contraindication_match: ap.contraindications.filter(ci => q.includes(ci)),
+        alternative: ap.alternative ? ARCHITECTURE_PATTERNS[ap.alternative]?.pattern : null,
+      });
+    } else {
+      detected.push({
+        anti_pattern: ap.name,
+        severity: ap.severity,
+        rationale: ap.rationale,
+        context_aware: key === 'over_engineering',
+        alternative: ap.alternative ? ARCHITECTURE_PATTERNS[ap.alternative]?.pattern : null,
+      });
+    }
+    detectedKeys.add(key);
+  }
+
+  // Check for over-engineering specifically (only if not already detected in loop)
+  if (!detectedKeys.has('over_engineering')) {
+    const overEng = ANTI_PATTERNS.over_engineering;
+    if (overEng.detection.test(q)) {
+      const hasJustification = overEng.contraindications.some(ci => q.includes(ci));
+      if (!hasJustification) {
+        detected.push({
+          anti_pattern: overEng.name,
+          severity: overEng.severity,
+          rationale: overEng.rationale,
+          context_aware: true,
+          missing_justification:
+            'No high-scale, regulatory, or multi-team requirement detected to justify this complexity.',
+          alternative: ARCHITECTURE_PATTERNS[overEng.alternative]?.pattern || null,
+        });
+      }
+    }
+  }
+
+  return detected;
+}
+
+// ============================================================================
+// 5. CANDIDATE GENERATION (SECTION 7)
+// ============================================================================
+
+/**
+ * Score how well a pattern matches requirements and constraints.
+ */
+function scoreCandidate(patternKey, pattern, requirements, constraints) {
+  let score = 0;
+  const reasons = [];
+  const risks = [];
+  const q = (requirements.raw_task || '').toLowerCase();
+  const allConstraints = (constraints && constraints.constraints) || [];
+
+  // Scale matching
+  const scaleConstraint = allConstraints.find(c => c.category === 'scale');
+  const scaleValue = scaleConstraint
+    ? parseInt(String(scaleConstraint.value).replace(/[,.\s]/g, ''), 10)
+    : 0;
+
+  if (scaleValue > 1000000) {
+    if (['microservices', 'event_driven', 'streaming'].includes(patternKey)) {
+      score += 3;
+      reasons.push('Architecture supports high-scale workloads');
+    } else if (['monolith'].includes(patternKey)) {
+      score -= 2;
+      risks.push(
+        'Single-process architecture may not handle millions of users without decomposition',
+      );
+    }
+  } else if (scaleValue > 0 && scaleValue <= 100000) {
+    if (['monolith', 'modular_monolith'].includes(patternKey)) {
+      score += 2;
+      reasons.push('Simpler architecture sufficient for moderate scale');
+    }
+    if (['microservices'].includes(patternKey)) {
+      score -= 1;
+      risks.push('Microservices complexity may not be justified at this scale');
+    }
+  }
+
+  // Team constraint
+  const teamConstraint = allConstraints.find(c => c.category === 'team');
+  if (teamConstraint && /(?:small|solo|1|2|3)/.test(String(teamConstraint.value))) {
+    if (['monolith', 'modular_monolith', 'serverless'].includes(patternKey)) {
+      score += 2;
+      reasons.push('Low operational overhead matches small team capacity');
+    }
+    if (['microservices', 'streaming'].includes(patternKey)) {
+      score -= 2;
+      risks.push('Small team cannot sustain operational overhead of distributed architecture');
+    }
+  }
+
+  // Financial domain
+  if (requirements.functional && requirements.functional.some(f => f.domain === 'financial')) {
+    if (['modular_monolith', 'hexagonal'].includes(patternKey)) {
+      score += 2;
+      reasons.push('Strong consistency and transaction support for financial operations');
+    }
+    if (['serverless'].includes(patternKey)) {
+      score -= 1;
+      risks.push(
+        'Serverless cold starts and execution limits may conflict with payment processing SLAs',
+      );
+    }
+  }
+
+  // Real-time / communication domain
+  if (
+    requirements.functional &&
+    requirements.functional.some(f => f.domain === 'communication' || f.domain === 'collaboration')
+  ) {
+    if (['event_driven', 'streaming'].includes(patternKey)) {
+      score += 2;
+      reasons.push('Event-driven architecture supports real-time communication patterns');
+    }
+  }
+
+  // Multi-tenancy
+  if (requirements.functional && requirements.functional.some(f => f.domain === 'isolation')) {
+    if (['modular_monolith', 'microservices'].includes(patternKey)) {
+      score += 1;
+      reasons.push('Architecture supports tenant isolation boundaries');
+    }
+  }
+
+  // Cost constraint
+  const costConstraint = allConstraints.find(c => c.category === 'cost');
+  if (costConstraint && /(?:low|limited|budget|startup|free)/i.test(String(costConstraint.value))) {
+    if (['monolith', 'modular_monolith', 'serverless'].includes(patternKey)) {
+      score += 1;
+      reasons.push('Lower infrastructure cost');
+    }
+    if (['microservices', 'streaming', 'event_sourcing'].includes(patternKey)) {
+      score -= 1;
+      risks.push('Higher infrastructure cost may exceed budget constraints');
+    }
+  }
+
+  // Timeline constraint
+  const timeConstraint = allConstraints.find(c => c.category === 'schedule');
+  if (
+    timeConstraint &&
+    /(?:mvp|prototype|poc|fast|quick|week)/i.test(String(timeConstraint.value || ''))
+  ) {
+    if (['monolith', 'modular_monolith', 'serverless'].includes(patternKey)) {
+      score += 2;
+      reasons.push('Faster time to first deployment');
+    }
+    if (['microservices', 'event_sourcing'].includes(patternKey)) {
+      score -= 1;
+      risks.push('Higher upfront investment delays time to market');
+    }
+  }
+
+  return { pattern: patternKey, name: pattern.pattern, score, reasons, risks };
+}
+
+/**
+ * Generate architecture candidates based on requirements and constraints.
+ *
+ * @param {Object} requirements
+ * @param {Object} constraints
+ * @param {Array} [considerations]
+ * @returns {Object} Architecture candidates with scores
+ */
+function generateCandidates(requirements, constraints, considerations) {
+  const scored = [];
+
+  for (const [key, pattern] of Object.entries(ARCHITECTURE_PATTERNS)) {
+    // Skip highly specific patterns that are sub-components rather than top-level architectures
+    if (['api_gateway', 'bff', 'saga', 'transactional_outbox'].includes(key)) continue;
+
+    const result = scoreCandidate(key, pattern, requirements, constraints);
+    scored.push(result);
+  }
+
+  // Sort by score descending
+  scored.sort((a, b) => b.score - a.score);
+
+  // Take top 2-4 candidates with meaningful differentiation
+  const topScore = scored[0]?.score || 0;
+  const candidates = scored.filter(s => s.score >= topScore - 3 && s.score >= -1).slice(0, 4);
+
+  return {
+    candidates: candidates.map((c, idx) => ({
+      rank: idx + 1,
+      pattern: c.pattern,
+      name: c.name,
+      score: c.score,
+      reasons: c.reasons,
+      risks: c.risks,
+      details: ARCHITECTURE_PATTERNS[c.pattern] || {},
+    })),
+    total_evaluated: scored.length,
+    recommendation: candidates[0] ? candidates[0].name : 'Unknown',
+    recommendation_rationale: candidates[0]
+      ? candidates[0].reasons.join('; ')
+      : 'No matching pattern found',
+  };
+}
+
+// ============================================================================
+// 6. TRADE-OFF ENGINE (SECTIONS 10–11)
+// ============================================================================
+
+const TRADEOFF_CRITERIA = [
+  'complexity',
+  'scalability',
+  'performance',
+  'reliability',
+  'availability',
+  'consistency',
+  'security',
+  'cost',
+  'development_speed',
+  'operational_burden',
+  'observability',
+  'testing_difficulty',
+  'deployment_complexity',
+  'recovery_complexity',
+  'team_complexity',
+  'data_management',
+  'vendor_lock_in',
+  'future_evolution',
+];
+
+/**
+ * Rate a candidate on a single criterion (LOW / MODERATE / HIGH / VERY_HIGH).
+ */
+function rateCriterion(patternKey, criterion) {
+  const ratings = {
+    monolith: {
+      complexity: 'LOW',
+      scalability: 'LOW',
+      performance: 'HIGH',
+      reliability: 'MODERATE',
+      availability: 'MODERATE',
+      consistency: 'HIGH',
+      security: 'MODERATE',
+      cost: 'LOW',
+      development_speed: 'HIGH',
+      operational_burden: 'LOW',
+      observability: 'MODERATE',
+      testing_difficulty: 'LOW',
+      deployment_complexity: 'LOW',
+      recovery_complexity: 'LOW',
+      team_complexity: 'LOW',
+      data_management: 'LOW',
+      vendor_lock_in: 'LOW',
+      future_evolution: 'LOW',
+    },
+    modular_monolith: {
+      complexity: 'MODERATE',
+      scalability: 'MODERATE',
+      performance: 'HIGH',
+      reliability: 'MODERATE',
+      availability: 'MODERATE',
+      consistency: 'HIGH',
+      security: 'MODERATE',
+      cost: 'LOW',
+      development_speed: 'HIGH',
+      operational_burden: 'LOW',
+      observability: 'MODERATE',
+      testing_difficulty: 'LOW',
+      deployment_complexity: 'LOW',
+      recovery_complexity: 'LOW',
+      team_complexity: 'LOW',
+      data_management: 'MODERATE',
+      vendor_lock_in: 'LOW',
+      future_evolution: 'MODERATE',
+    },
+    microservices: {
+      complexity: 'VERY_HIGH',
+      scalability: 'VERY_HIGH',
+      performance: 'MODERATE',
+      reliability: 'HIGH',
+      availability: 'HIGH',
+      consistency: 'LOW',
+      security: 'HIGH',
+      cost: 'HIGH',
+      development_speed: 'LOW',
+      operational_burden: 'VERY_HIGH',
+      observability: 'HIGH',
+      testing_difficulty: 'HIGH',
+      deployment_complexity: 'HIGH',
+      recovery_complexity: 'MODERATE',
+      team_complexity: 'HIGH',
+      data_management: 'HIGH',
+      vendor_lock_in: 'MODERATE',
+      future_evolution: 'HIGH',
+    },
+    serverless: {
+      complexity: 'MODERATE',
+      scalability: 'HIGH',
+      performance: 'MODERATE',
+      reliability: 'HIGH',
+      availability: 'HIGH',
+      consistency: 'MODERATE',
+      security: 'MODERATE',
+      cost: 'LOW',
+      development_speed: 'HIGH',
+      operational_burden: 'LOW',
+      observability: 'MODERATE',
+      testing_difficulty: 'MODERATE',
+      deployment_complexity: 'LOW',
+      recovery_complexity: 'LOW',
+      team_complexity: 'LOW',
+      data_management: 'MODERATE',
+      vendor_lock_in: 'HIGH',
+      future_evolution: 'MODERATE',
+    },
+    event_driven: {
+      complexity: 'HIGH',
+      scalability: 'HIGH',
+      performance: 'MODERATE',
+      reliability: 'HIGH',
+      availability: 'HIGH',
+      consistency: 'LOW',
+      security: 'MODERATE',
+      cost: 'MODERATE',
+      development_speed: 'MODERATE',
+      operational_burden: 'MODERATE',
+      observability: 'MODERATE',
+      testing_difficulty: 'HIGH',
+      deployment_complexity: 'MODERATE',
+      recovery_complexity: 'MODERATE',
+      team_complexity: 'MODERATE',
+      data_management: 'MODERATE',
+      vendor_lock_in: 'MODERATE',
+      future_evolution: 'HIGH',
+    },
+    cqrs: {
+      complexity: 'HIGH',
+      scalability: 'HIGH',
+      performance: 'HIGH',
+      reliability: 'MODERATE',
+      availability: 'MODERATE',
+      consistency: 'LOW',
+      security: 'MODERATE',
+      cost: 'MODERATE',
+      development_speed: 'LOW',
+      operational_burden: 'MODERATE',
+      observability: 'MODERATE',
+      testing_difficulty: 'HIGH',
+      deployment_complexity: 'MODERATE',
+      recovery_complexity: 'MODERATE',
+      team_complexity: 'MODERATE',
+      data_management: 'HIGH',
+      vendor_lock_in: 'LOW',
+      future_evolution: 'HIGH',
+    },
+    streaming: {
+      complexity: 'VERY_HIGH',
+      scalability: 'VERY_HIGH',
+      performance: 'HIGH',
+      reliability: 'HIGH',
+      availability: 'HIGH',
+      consistency: 'LOW',
+      security: 'MODERATE',
+      cost: 'HIGH',
+      development_speed: 'LOW',
+      operational_burden: 'HIGH',
+      observability: 'MODERATE',
+      testing_difficulty: 'HIGH',
+      deployment_complexity: 'HIGH',
+      recovery_complexity: 'MODERATE',
+      team_complexity: 'HIGH',
+      data_management: 'HIGH',
+      vendor_lock_in: 'MODERATE',
+      future_evolution: 'HIGH',
+    },
+    hexagonal: {
+      complexity: 'MODERATE',
+      scalability: 'MODERATE',
+      performance: 'HIGH',
+      reliability: 'MODERATE',
+      availability: 'MODERATE',
+      consistency: 'HIGH',
+      security: 'MODERATE',
+      cost: 'LOW',
+      development_speed: 'MODERATE',
+      operational_burden: 'LOW',
+      observability: 'MODERATE',
+      testing_difficulty: 'LOW',
+      deployment_complexity: 'LOW',
+      recovery_complexity: 'LOW',
+      team_complexity: 'MODERATE',
+      data_management: 'LOW',
+      vendor_lock_in: 'LOW',
+      future_evolution: 'HIGH',
+    },
+    batch: {
+      complexity: 'LOW',
+      scalability: 'MODERATE',
+      performance: 'LOW',
+      reliability: 'MODERATE',
+      availability: 'LOW',
+      consistency: 'HIGH',
+      security: 'MODERATE',
+      cost: 'LOW',
+      development_speed: 'HIGH',
+      operational_burden: 'LOW',
+      observability: 'LOW',
+      testing_difficulty: 'LOW',
+      deployment_complexity: 'LOW',
+      recovery_complexity: 'LOW',
+      team_complexity: 'LOW',
+      data_management: 'MODERATE',
+      vendor_lock_in: 'LOW',
+      future_evolution: 'LOW',
+    },
+    distributed_workers: {
+      complexity: 'MODERATE',
+      scalability: 'HIGH',
+      performance: 'MODERATE',
+      reliability: 'HIGH',
+      availability: 'MODERATE',
+      consistency: 'MODERATE',
+      security: 'MODERATE',
+      cost: 'MODERATE',
+      development_speed: 'MODERATE',
+      operational_burden: 'MODERATE',
+      observability: 'MODERATE',
+      testing_difficulty: 'MODERATE',
+      deployment_complexity: 'MODERATE',
+      recovery_complexity: 'MODERATE',
+      team_complexity: 'MODERATE',
+      data_management: 'MODERATE',
+      vendor_lock_in: 'LOW',
+      future_evolution: 'MODERATE',
+    },
+    event_sourcing: {
+      complexity: 'VERY_HIGH',
+      scalability: 'HIGH',
+      performance: 'MODERATE',
+      reliability: 'HIGH',
+      availability: 'MODERATE',
+      consistency: 'HIGH',
+      security: 'MODERATE',
+      cost: 'HIGH',
+      development_speed: 'LOW',
+      operational_burden: 'HIGH',
+      observability: 'HIGH',
+      testing_difficulty: 'HIGH',
+      deployment_complexity: 'MODERATE',
+      recovery_complexity: 'MODERATE',
+      team_complexity: 'HIGH',
+      data_management: 'VERY_HIGH',
+      vendor_lock_in: 'LOW',
+      future_evolution: 'HIGH',
+    },
+  };
+
+  const patternRatings = ratings[patternKey];
+  if (!patternRatings) return 'UNKNOWN';
+  return patternRatings[criterion] || 'UNKNOWN';
+}
+
+/**
+ * Build a full trade-off matrix comparing candidates.
+ *
+ * @param {Array} candidates
+ * @param {Object} requirements
+ * @param {Object} constraints
+ * @returns {Object} Trade-off matrix
+ */
+function analyzeTradeoffs(candidates, requirements, constraints) {
+  if (!candidates || candidates.length < 2) {
+    return {
+      matrix: [],
+      summary: 'Single candidate — no trade-off comparison needed.',
+      meaningful_differences: [],
+    };
+  }
+
+  const matrix = [];
+  const meaningfulDifferences = [];
+
+  for (const criterion of TRADEOFF_CRITERIA) {
+    const row = {
+      criterion,
+      importance: determineCriterionImportance(criterion, requirements, constraints),
+      candidate_values: {},
+      evidence: [],
+      uncertainty: 'MODERATE',
+    };
+
+    const values = {};
+    for (const cand of candidates) {
+      const rating = rateCriterion(cand.pattern, criterion);
+      row.candidate_values[cand.name] = rating;
+      values[cand.name] = rating;
+    }
+
+    // Detect meaningful differences
+    const uniqueValues = new Set(Object.values(values));
+    if (uniqueValues.size > 1 && row.importance !== 'LOW') {
+      meaningfulDifferences.push({
+        criterion,
+        importance: row.importance,
+        values,
+      });
+    }
+
+    matrix.push(row);
+  }
+
+  return {
+    matrix,
+    summary: `Compared ${candidates.length} candidates across ${TRADEOFF_CRITERIA.length} criteria. ${meaningfulDifferences.length} meaningful differences found.`,
+    meaningful_differences: meaningfulDifferences,
+    candidates_compared: candidates.map(c => c.name),
+  };
+}
+
+/**
+ * Determine how important a criterion is given the requirements.
+ */
+function determineCriterionImportance(criterion, requirements, constraints) {
+  const nfrs = (requirements && requirements.non_functional) || [];
+  const frs = (requirements && requirements.functional) || [];
+
+  const highImportance = {
+    scalability: nfrs.some(n => n.category === 'scalability'),
+    performance: nfrs.some(n => n.category === 'performance'),
+    reliability: nfrs.some(n => n.category === 'reliability'),
+    security:
+      nfrs.some(n => n.category === 'security') ||
+      frs.some(f => f.domain === 'financial' || f.domain === 'security'),
+    consistency: frs.some(f => f.domain === 'financial'),
+    cost: ((constraints && constraints.constraints) || []).some(c => c.category === 'cost'),
+    operational_burden: ((constraints && constraints.constraints) || []).some(
+      c => c.category === 'team' || c.category === 'operations',
+    ),
+    availability: nfrs.some(n => n.category === 'reliability'),
+  };
+
+  if (highImportance[criterion]) return 'HIGH';
+  if (['complexity', 'development_speed', 'testing_difficulty'].includes(criterion))
+    return 'MEDIUM';
+  return 'LOW';
+}
+
+// ============================================================================
+// 7. DECISION SENSITIVITY & BREAKPOINTS (SECTIONS 12–13)
+// ============================================================================
+
+/**
+ * Identify which assumptions could change the architecture decision.
+ *
+ * @param {Array} candidates
+ * @param {Object} tradeoffs
+ * @param {Object} requirements
+ * @returns {Array<Object>} Decision-sensitive assumptions
+ */
+function analyzeDecisionSensitivity(candidates, tradeoffs, requirements) {
+  const sensitivities = [];
+  const unknowns = (requirements && requirements.unknowns) || [];
+
+  // Traffic assumption sensitivity
+  if (
+    unknowns.some(
+      u =>
+        u.requirement === 'request rate / throughput' || u.requirement === 'peak concurrent users',
+    )
+  ) {
+    sensitivities.push({
+      assumption: 'Current traffic estimate',
+      current_value: 'UNKNOWN',
+      threshold: 'If traffic exceeds 100K requests/day, consider service decomposition.',
+      impact: 'Architecture decision may shift from monolith to service-based.',
+      type: 'DECISION_SENSITIVE',
+      evidence_quality: 'UNKNOWN',
+    });
+  }
+
+  // Team growth sensitivity
+  if (unknowns.some(u => u.requirement === 'team size and expertise')) {
+    sensitivities.push({
+      assumption: 'Team size remains stable',
+      current_value: 'UNKNOWN',
+      threshold:
+        'If team grows beyond 8-10 engineers, modular monolith may need service decomposition.',
+      impact: 'Architecture decision may shift to support independent team ownership.',
+      type: 'DECISION_SENSITIVE',
+      evidence_quality: 'UNKNOWN',
+    });
+  }
+
+  // Data volume sensitivity
+  if (unknowns.some(u => u.requirement === 'data retention period')) {
+    sensitivities.push({
+      assumption: 'Data volume remains manageable',
+      current_value: 'UNKNOWN',
+      threshold: 'If data exceeds single-node DB capacity, partitioning/sharding strategy needed.',
+      impact: 'Database architecture may need fundamental change.',
+      type: 'DECISION_SENSITIVE',
+      evidence_quality: 'UNKNOWN',
+    });
+  }
+
+  // Cost sensitivity
+  if (unknowns.some(u => u.requirement === 'budget')) {
+    sensitivities.push({
+      assumption: 'Budget is sufficient for chosen architecture',
+      current_value: 'UNKNOWN',
+      threshold:
+        'If budget is severely constrained, prefer monolith/serverless over microservices.',
+      impact: 'Cost model may invalidate complex architecture choices.',
+      type: 'DECISION_SENSITIVE',
+      evidence_quality: 'UNKNOWN',
+    });
+  }
+
+  return sensitivities;
+}
+
+/**
+ * Identify architectural breakpoints — thresholds where architecture should change.
+ *
+ * @param {Object} candidate
+ * @param {Object} requirements
+ * @returns {Array<Object>} Breakpoints
+ */
+function identifyBreakpoints(candidate, requirements) {
+  const breakpoints = [];
+  const patternKey = candidate.pattern || '';
+
+  if (['monolith', 'modular_monolith'].includes(patternKey)) {
+    breakpoints.push({
+      dimension: 'traffic',
+      current_range: '0–100K requests/day',
+      breakpoint: '100K–1M requests/day',
+      action: 'Scale application horizontally + database read replicas',
+      evidence_quality: 'ESTIMATED',
+      next_architecture: 'Consider service decomposition at >1M requests/day',
+    });
+    breakpoints.push({
+      dimension: 'team_size',
+      current_range: '1–8 engineers',
+      breakpoint: '8–15 engineers',
+      action: 'Enforce module ownership boundaries, consider domain service extraction',
+      evidence_quality: 'ESTIMATED',
+      next_architecture: 'Module → Service transition for >15 engineers',
+    });
+    breakpoints.push({
+      dimension: 'data_volume',
+      current_range: 'Single-node database',
+      breakpoint: 'Database exceeds single-node capacity',
+      action: 'Implement read replicas, then partitioning/sharding',
+      evidence_quality: 'ESTIMATED',
+      next_architecture: 'Per-domain database ownership',
+    });
+  }
+
+  if (['microservices'].includes(patternKey)) {
+    breakpoints.push({
+      dimension: 'team_size',
+      current_range: '>8 engineers',
+      breakpoint: '<4 engineers',
+      action:
+        'Consolidate services into modular monolith — team cannot sustain operational overhead',
+      evidence_quality: 'ESTIMATED',
+      next_architecture: 'Modular Monolith',
+    });
+  }
+
+  if (['serverless'].includes(patternKey)) {
+    breakpoints.push({
+      dimension: 'execution_duration',
+      current_range: 'Sub-15-minute tasks',
+      breakpoint: 'Tasks exceeding runtime limits',
+      action: 'Move long-running workflows to container-based workers',
+      evidence_quality: 'DOCUMENTED',
+      next_architecture: 'Distributed Workers or container-based services',
+    });
+    breakpoints.push({
+      dimension: 'cold_start_tolerance',
+      current_range: 'Acceptable cold start latency',
+      breakpoint: 'p99 latency exceeds SLA due to cold starts',
+      action: 'Provisioned concurrency or move to always-on containers',
+      evidence_quality: 'MEASURED',
+      next_architecture: 'Container-based deployment',
+    });
+  }
+
+  return breakpoints;
+}
+
+// ============================================================================
+// 8. DOMAIN-SPECIFIC ANALYZERS (SECTIONS 14–27)
+// ============================================================================
+
+/**
+ * Analyze build vs buy decisions for common components.
+ */
+function analyzeBuildVsBuy(taskQuery, requirements) {
+  const q = (taskQuery || '').toLowerCase();
+  const decisions = [];
+
+  const components = [
+    {
+      match: /(?:auth|login|sso|oauth)/i,
+      component: 'Authentication',
+      buy: 'Auth0, Clerk, Supabase Auth, Firebase Auth',
+      build_reason: 'Custom auth flows, data sovereignty, no vendor dependency',
+      buy_reason: 'Faster implementation, proven security, compliance certifications',
+    },
+    {
+      match: /(?:payment|billing|subscription|charge)/i,
+      component: 'Payment Processing',
+      buy: 'Stripe, Paddle, LemonSqueezy',
+      build_reason: 'Rarely justified — regulatory and PCI compliance burden',
+      buy_reason: 'PCI compliance handled, fraud detection, global payment methods',
+    },
+    {
+      match: /(?:search|full.text|elastic)/i,
+      component: 'Search',
+      buy: 'Algolia, Meilisearch, Typesense, ElasticSearch (managed)',
+      build_reason: 'Custom ranking algorithms, data privacy',
+      buy_reason: 'Performance, relevance tuning, faceted search out-of-box',
+    },
+    {
+      match: /(?:email|smtp|transactional.mail)/i,
+      component: 'Email',
+      buy: 'Resend, SendGrid, Postmark, AWS SES',
+      build_reason: 'Almost never justified',
+      buy_reason: 'Deliverability, reputation management, compliance',
+    },
+    {
+      match: /(?:observab|monitor|trace|log|metric)/i,
+      component: 'Observability',
+      buy: 'Datadog, Grafana Cloud, New Relic',
+      build_reason: 'Cost at scale, data privacy',
+      buy_reason: 'Immediate visibility, alerting, dashboards',
+    },
+    {
+      match: /(?:vector|embedding|rag|semantic.search)/i,
+      component: 'Vector Database',
+      buy: 'Pinecone, Weaviate Cloud, Qdrant Cloud',
+      build_reason: 'pgvector for simple cases, data sovereignty',
+      buy_reason: 'Optimized ANN search, managed scaling',
+    },
+    {
+      match: /(?:feature.flag|toggle|experiment)/i,
+      component: 'Feature Flags',
+      buy: 'LaunchDarkly, Flagsmith, Unleash',
+      build_reason: 'Simple boolean flags can be DIY',
+      buy_reason: 'Percentage rollouts, A/B testing, audit trail',
+    },
+  ];
+
+  for (const c of components) {
+    if (c.match.test(q)) {
+      decisions.push({
+        component: c.component,
+        recommendation:
+          c.component === 'Payment Processing' || c.component === 'Email' ? 'BUY' : 'EVALUATE',
+        buy_options: c.buy,
+        build_rationale: c.build_reason,
+        buy_rationale: c.buy_reason,
+      });
+    }
+  }
+
+  return decisions;
+}
+
+/**
+ * Analyze data architecture requirements.
+ */
+function analyzeDataArchitecture(taskQuery, requirements) {
+  const q = (taskQuery || '').toLowerCase();
+  const analysis = {
+    consistency_model: 'UNKNOWN',
+    transaction_boundaries: [],
+    read_write_pattern: 'UNKNOWN',
+    recommendations: [],
+    considerations: [],
+  };
+
+  // Consistency model
+  if (/(?:payment|financial|ledger|order|inventory|booking)/i.test(q)) {
+    analysis.consistency_model = 'STRONG';
+    analysis.considerations.push(
+      'Financial data requires strong consistency — use ACID transactions.',
+    );
+    analysis.transaction_boundaries.push(
+      'payment execution',
+      'inventory reservation',
+      'order creation',
+    );
+  } else if (/(?:feed|timeline|analytics|recommendation|cache)/i.test(q)) {
+    analysis.consistency_model = 'EVENTUAL';
+    analysis.considerations.push(
+      'Feed/analytics data can tolerate eventual consistency for better performance.',
+    );
+  } else if (/(?:collaboration|editor|realtime)/i.test(q)) {
+    analysis.consistency_model = 'CAUSAL';
+    analysis.considerations.push(
+      'Collaborative editing requires causal consistency with conflict resolution (CRDT/OT).',
+    );
+  }
+
+  // Read/write pattern
+  if (/(?:feed|timeline|dashboard|report|analytics)/i.test(q)) {
+    analysis.read_write_pattern = 'READ_HEAVY';
+    analysis.recommendations.push('Consider read replicas, caching, and materialized views.');
+  } else if (/(?:log|audit|event|ingest|import)/i.test(q)) {
+    analysis.read_write_pattern = 'WRITE_HEAVY';
+    analysis.recommendations.push(
+      'Consider write-optimized storage, append-only logs, and batch inserts.',
+    );
+  } else {
+    analysis.read_write_pattern = 'BALANCED';
+  }
+
+  // Database recommendations
+  if (
+    /(?:relational|sql|postgres|mysql|transaction|acid|join|foreign.key)/i.test(q) ||
+    analysis.consistency_model === 'STRONG'
+  ) {
+    analysis.recommendations.push(
+      'PostgreSQL: Strong ACID, extensible (pgvector, PostGIS), mature ecosystem.',
+    );
+  }
+  if (/(?:document|json|flexible.schema|mongo|nosql)/i.test(q)) {
+    analysis.recommendations.push(
+      'Document store (MongoDB, DynamoDB): Flexible schema, horizontal scaling. Evaluate if relational model is truly insufficient.',
+    );
+  }
+  if (/(?:time.series|iot|sensor|metric|tsdb)/i.test(q)) {
+    analysis.recommendations.push(
+      'Time-series DB (TimescaleDB, InfluxDB): Optimized for temporal queries and retention.',
+    );
+  }
+  if (/(?:graph|relation|connection|social|network|knowledge)/i.test(q)) {
+    analysis.recommendations.push(
+      'Graph DB (Neo4j, Neptune): Consider if relationships are the primary query target.',
+    );
+  }
+
+  return analysis;
+}
+
+/**
+ * Analyze communication patterns.
+ */
+function analyzeCommunicationPatterns(taskQuery) {
+  const q = (taskQuery || '').toLowerCase();
+  const patterns = [];
+
+  if (/(?:api|endpoint|crud|rest)/i.test(q)) {
+    patterns.push({
+      pattern: 'REST',
+      rationale: 'Standard HTTP API for CRUD operations',
+      fit: 'HIGH',
+    });
+  }
+  if (/(?:realtime|websocket|presence|chat|collaboration)/i.test(q)) {
+    patterns.push({
+      pattern: 'WebSocket',
+      rationale: 'Bidirectional real-time communication',
+      fit: 'HIGH',
+    });
+  }
+  if (/(?:notification|webhook|event.trigger)/i.test(q)) {
+    patterns.push({ pattern: 'Webhook', rationale: 'Push-based event notification', fit: 'HIGH' });
+  }
+  if (/(?:stream|sse|server.sent)/i.test(q)) {
+    patterns.push({ pattern: 'SSE', rationale: 'Server-to-client streaming', fit: 'MODERATE' });
+  }
+  if (/(?:grpc|protobuf|internal.service|high.throughput.rpc)/i.test(q)) {
+    patterns.push({
+      pattern: 'gRPC',
+      rationale: 'High-performance service-to-service communication',
+      fit: 'MODERATE',
+    });
+  }
+  if (/(?:graphql|flexible.query|client.driven)/i.test(q)) {
+    patterns.push({
+      pattern: 'GraphQL',
+      rationale: 'Client-driven flexible queries',
+      fit: 'MODERATE',
+    });
+  }
+  if (/(?:queue|async|background|worker|job)/i.test(q)) {
+    patterns.push({
+      pattern: 'Message Queue',
+      rationale: 'Asynchronous decoupled processing',
+      fit: 'HIGH',
+    });
+  }
+
+  if (patterns.length === 0) {
+    patterns.push({
+      pattern: 'REST',
+      rationale: 'Default API pattern for HTTP services',
+      fit: 'MODERATE',
+    });
+  }
+
+  return patterns;
+}
+
+/**
+ * Analyze security architecture.
+ */
+function analyzeSecurityArchitecture(taskQuery, requirements) {
+  const q = (taskQuery || '').toLowerCase();
+  const analysis = {
+    trust_boundaries: [],
+    authentication_model: 'UNKNOWN',
+    authorization_model: 'UNKNOWN',
+    considerations: [],
+    trust_boundary_graph: [],
+  };
+
+  // Trust boundary graph
+  analysis.trust_boundary_graph = [
+    { from: 'Internet', to: 'CDN/WAF', trust_transition: true },
+    { from: 'CDN/WAF', to: 'Load Balancer', trust_transition: false },
+    { from: 'Load Balancer', to: 'Application', trust_transition: true },
+    { from: 'Application', to: 'Database', trust_transition: true },
+  ];
+
+  if (/(?:auth|login|jwt|session|oauth|sso)/i.test(q)) {
+    analysis.authentication_model = /(?:jwt|token)/i.test(q) ? 'JWT/Token-based' : 'Session-based';
+    analysis.considerations.push('Enforce cryptographic signature verification on all tokens.');
+    analysis.considerations.push('Implement token expiration and refresh rotation.');
+  }
+
+  if (/(?:rbac|role|permission|acl|authorization)/i.test(q)) {
+    analysis.authorization_model = 'RBAC';
+    analysis.considerations.push('Define authorization matrix: User × Role × Resource × Action.');
+    analysis.considerations.push('Enforce authorization before business logic, never after.');
+  }
+
+  if (/(?:multi.tenant|tenant|workspace)/i.test(q)) {
+    analysis.considerations.push('CRITICAL: Tenant isolation at every data access boundary.');
+    analysis.considerations.push(
+      'Verify cross-tenant access is impossible at database query level.',
+    );
+    analysis.trust_boundary_graph.push({
+      from: 'Application',
+      to: 'Tenant Boundary',
+      trust_transition: true,
+    });
+  }
+
+  if (/(?:agent|llm|prompt|ai|tool)/i.test(q)) {
+    analysis.considerations.push(
+      'Agent sandboxing: restrict subprocess execution and tool permissions.',
+    );
+    analysis.considerations.push(
+      'Prompt injection defense: isolate user input from system instructions.',
+    );
+    analysis.trust_boundary_graph.push({
+      from: 'Application',
+      to: 'AI Agent Sandbox',
+      trust_transition: true,
+    });
+  }
+
+  return analysis;
+}
+
+/**
+ * Analyze cost architecture for a candidate.
+ */
+function analyzeCostArchitecture(candidate, constraints) {
+  const patternKey = candidate.pattern || '';
+  const costDrivers = {
+    compute: 'UNKNOWN',
+    database: 'UNKNOWN',
+    storage: 'UNKNOWN',
+    network: 'UNKNOWN',
+    managed_services: 'UNKNOWN',
+    observability: 'UNKNOWN',
+    operations: 'UNKNOWN',
+  };
+
+  const costProfiles = {
+    monolith: {
+      compute: 'LOW',
+      database: 'LOW',
+      storage: 'LOW',
+      network: 'LOW',
+      managed_services: 'LOW',
+      observability: 'LOW',
+      operations: 'LOW',
+    },
+    modular_monolith: {
+      compute: 'LOW',
+      database: 'LOW',
+      storage: 'LOW',
+      network: 'LOW',
+      managed_services: 'LOW',
+      observability: 'LOW',
+      operations: 'LOW',
+    },
+    microservices: {
+      compute: 'HIGH',
+      database: 'HIGH',
+      storage: 'MODERATE',
+      network: 'HIGH',
+      managed_services: 'HIGH',
+      observability: 'HIGH',
+      operations: 'HIGH',
+    },
+    serverless: {
+      compute: 'VARIABLE',
+      database: 'MODERATE',
+      storage: 'LOW',
+      network: 'LOW',
+      managed_services: 'MODERATE',
+      observability: 'MODERATE',
+      operations: 'LOW',
+    },
+    event_driven: {
+      compute: 'MODERATE',
+      database: 'MODERATE',
+      storage: 'MODERATE',
+      network: 'MODERATE',
+      managed_services: 'MODERATE',
+      observability: 'MODERATE',
+      operations: 'MODERATE',
+    },
+    streaming: {
+      compute: 'HIGH',
+      database: 'HIGH',
+      storage: 'HIGH',
+      network: 'HIGH',
+      managed_services: 'HIGH',
+      observability: 'HIGH',
+      operations: 'HIGH',
+    },
+  };
+
+  const profile = costProfiles[patternKey] || {};
+  Object.assign(costDrivers, profile);
+
+  return {
+    pattern: candidate.name || patternKey,
+    cost_drivers: costDrivers,
+    cost_category: ['monolith', 'modular_monolith'].includes(patternKey)
+      ? 'LOW'
+      : ['microservices', 'streaming'].includes(patternKey)
+        ? 'HIGH'
+        : 'MODERATE',
+    scaling_cost_model: ['serverless'].includes(patternKey) ? 'PAY_PER_USE' : 'FIXED_PLUS_VARIABLE',
+    notes: [],
+  };
+}
+
+/**
+ * Analyze operational complexity for a candidate.
+ */
+function analyzeOperationalComplexity(candidate) {
+  const patternKey = candidate.pattern || '';
+  const profiles = {
+    monolith: {
+      deployment_units: 1,
+      services: 1,
+      databases: 1,
+      queues: 0,
+      secrets: 'FEW',
+      monitoring: 'BASIC',
+      on_call: 'LOW',
+    },
+    modular_monolith: {
+      deployment_units: 1,
+      services: 1,
+      databases: 1,
+      queues: 0,
+      secrets: 'FEW',
+      monitoring: 'BASIC',
+      on_call: 'LOW',
+    },
+    microservices: {
+      deployment_units: 'MANY',
+      services: 'MANY',
+      databases: 'MANY',
+      queues: 'SEVERAL',
+      secrets: 'MANY',
+      monitoring: 'COMPLEX',
+      on_call: 'HIGH',
+    },
+    serverless: {
+      deployment_units: 'MANY_FUNCTIONS',
+      services: 'MANY_FUNCTIONS',
+      databases: 1,
+      queues: 'FEW',
+      secrets: 'MODERATE',
+      monitoring: 'MODERATE',
+      on_call: 'LOW',
+    },
+    event_driven: {
+      deployment_units: 'SEVERAL',
+      services: 'SEVERAL',
+      databases: 'FEW',
+      queues: 'SEVERAL',
+      secrets: 'MODERATE',
+      monitoring: 'MODERATE',
+      on_call: 'MODERATE',
+    },
+    streaming: {
+      deployment_units: 'MANY',
+      services: 'MANY',
+      databases: 'SEVERAL',
+      queues: 'MANY',
+      secrets: 'MANY',
+      monitoring: 'COMPLEX',
+      on_call: 'HIGH',
+    },
+  };
+
+  return (
+    profiles[patternKey] || {
+      deployment_units: 'UNKNOWN',
+      services: 'UNKNOWN',
+      databases: 'UNKNOWN',
+      queues: 'UNKNOWN',
+      secrets: 'UNKNOWN',
+      monitoring: 'UNKNOWN',
+      on_call: 'UNKNOWN',
+    }
+  );
+}
+
+// ============================================================================
+// 9. MIGRATION & EVOLUTION INTELLIGENCE (SECTIONS 28–29)
+// ============================================================================
+
+/**
+ * Analyze migration path between architectures.
+ */
+function analyzeMigration(currentArch, targetArch) {
+  const migration = {
+    current: currentArch,
+    target: targetArch,
+    strategy: 'INCREMENTAL',
+    risk_level: 'MODERATE',
+    steps: [],
+    rollback_plan: 'Maintain dual deployment during migration; route traffic via feature flags.',
+    considerations: [],
+  };
+
+  if (currentArch === 'monolith' && targetArch === 'modular_monolith') {
+    migration.risk_level = 'LOW';
+    migration.steps = [
+      'Identify domain boundaries in existing code',
+      'Extract shared code into internal library modules',
+      'Define module interfaces (internal API contracts)',
+      'Enforce module boundaries via import restrictions',
+      'Add module-level integration tests',
+    ];
+  } else if (currentArch === 'modular_monolith' && targetArch === 'microservices') {
+    migration.risk_level = 'HIGH';
+    migration.strategy = 'STRANGLER_FIG';
+    migration.steps = [
+      'Identify highest-value module for extraction (highest deployment friction)',
+      'Set up service infrastructure (discovery, auth, observability)',
+      'Extract module to independent service behind API gateway',
+      'Implement dual-write/read with consistency verification',
+      'Migrate traffic incrementally via feature flags',
+      'Decommission monolith module after verification period',
+    ];
+    migration.considerations.push(
+      'CRITICAL: Big-bang migration has high failure risk. Use Strangler Fig pattern.',
+    );
+  }
+
+  return migration;
+}
+
+/**
+ * Analyze evolution paths for an architecture.
+ */
+function analyzeEvolution(candidate) {
+  const patternKey = candidate.pattern || '';
+  const evolution = {
+    pattern: candidate.name || patternKey,
+    evolution_paths: [],
+    growth_directions: [],
+  };
+
+  if (['monolith', 'modular_monolith'].includes(patternKey)) {
+    evolution.evolution_paths = [
+      {
+        trigger: 'Traffic growth beyond single-node',
+        target: 'Horizontal scaling + read replicas',
+      },
+      { trigger: 'Team growth beyond 8–10', target: 'Service decomposition via Strangler Fig' },
+      {
+        trigger: 'New compliance requirements',
+        target: 'Add audit logging module, encryption at rest',
+      },
+      { trigger: 'Geographic expansion', target: 'Multi-region deployment with read replicas' },
+    ];
+    evolution.growth_directions = ['vertical_first', 'then_horizontal', 'then_decomposition'];
+  }
+
+  if (['microservices'].includes(patternKey)) {
+    evolution.evolution_paths = [
+      { trigger: 'Team shrinkage', target: 'Consolidate services into fewer, larger services' },
+      {
+        trigger: 'Service mesh complexity',
+        target: 'Evaluate managed service mesh or simplification',
+      },
+      {
+        trigger: 'Data consistency challenges',
+        target: 'Introduce saga orchestrator or event sourcing',
+      },
+    ];
+    evolution.growth_directions = ['service_per_team', 'independent_scaling'];
+  }
+
+  if (['serverless'].includes(patternKey)) {
+    evolution.evolution_paths = [
+      {
+        trigger: 'Cold start exceeds SLA',
+        target: 'Provisioned concurrency or container migration',
+      },
+      {
+        trigger: 'Vendor lock-in concern',
+        target: 'Abstract via ports & adapters for portability',
+      },
+      {
+        trigger: 'Long-running workflows',
+        target: 'Step Functions / Durable Functions or container workers',
+      },
+    ];
+    evolution.growth_directions = ['function_per_feature', 'managed_services'];
+  }
+
+  return evolution;
+}
+
+// ============================================================================
+// 10. ADR GENERATION (SECTION 30)
+// ============================================================================
+
+/**
+ * Generate an Architecture Decision Record.
+ *
+ * @param {Object} params
+ * @returns {Object} ADR
+ */
+function generateADR(params) {
+  const {
+    task,
+    requirements,
+    constraints,
+    candidates,
+    tradeoffs,
+    decision,
+    antiPatterns,
+    sensitivity,
+    breakpoints,
+    securityAnalysis,
+  } = params;
+
+  const selectedCandidate = decision || (candidates && candidates[0]) || {};
+  const rejectedCandidates = (candidates || []).filter(c => c.name !== selectedCandidate.name);
+
+  return {
+    id: `ADR-${crypto.randomUUID ? crypto.randomUUID().slice(0, 8) : Date.now().toString(36)}`,
+    title: `Architecture Decision: ${selectedCandidate.name || 'Unknown'}`,
+    status: 'PROPOSED',
+    date: new Date().toISOString(),
+    context: {
+      task: task || '',
+      functional_requirements: (requirements && requirements.functional) || [],
+      non_functional_requirements: (requirements && requirements.non_functional) || [],
+      constraints: (constraints && constraints.constraints) || [],
+      unknowns: (requirements && requirements.unknowns) || [],
+    },
+    decision: {
+      architecture: selectedCandidate.name || 'Unknown',
+      pattern: selectedCandidate.pattern || '',
+      rationale: selectedCandidate.reasons || [],
+      score: selectedCandidate.score || 0,
+    },
+    tradeoffs: tradeoffs || {},
+    rejected_alternatives: rejectedCandidates.map(c => ({
+      architecture: c.name,
+      pattern: c.pattern,
+      rejection_reason: c.risks || [],
+      score: c.score || 0,
+    })),
+    anti_patterns_detected: antiPatterns || [],
+    assumptions: (requirements && requirements.assumptions) || [],
+    evidence: [],
+    risks: selectedCandidate.risks || [],
+    decision_sensitivity: sensitivity || [],
+    breakpoints: breakpoints || [],
+    security_considerations: (securityAnalysis && securityAnalysis.considerations) || [],
+    verification_plan: [
+      'Verify architecture handles expected load via load test',
+      'Verify failure behavior via Phase 6 pre-mortem simulation',
+      'Verify security boundaries via Phase 8 validation',
+      'Review decision sensitivity assumptions quarterly',
+    ],
+    revisit_conditions: [
+      'Traffic exceeds estimated breakpoint thresholds',
+      'Team size changes significantly',
+      'New compliance requirements imposed',
+      'Cost exceeds projected budget by >50%',
+      'Critical failure mode materializes in production',
+    ],
+  };
+}
+
+// ============================================================================
+// 11. ARCHITECTURE GRAPH (SECTION 39)
+// ============================================================================
+
+/**
+ * Generate a machine-readable architecture graph.
+ */
+function generateArchitectureGraph(candidate, requirements, securityAnalysis) {
+  const patternKey = candidate.pattern || '';
+  const nodes = [];
+  const edges = [];
+
+  // Always have a client
+  nodes.push({
+    id: 'client',
+    type: 'external',
+    technology: 'Browser/Mobile',
+    responsibility: 'User interaction',
+    scaling_model: 'N/A',
+  });
+
+  // API layer
+  nodes.push({
+    id: 'api',
+    type: 'service',
+    technology: 'API Server',
+    responsibility: 'Request handling, routing, validation',
+    scaling_model: 'horizontal',
+    security_boundary: true,
+  });
+  edges.push({ from: 'client', to: 'api', protocol: 'HTTPS' });
+
+  // Database
+  nodes.push({
+    id: 'database',
+    type: 'datastore',
+    technology: 'PostgreSQL',
+    responsibility: 'Primary data persistence',
+    scaling_model: 'vertical + read replicas',
+    security_boundary: true,
+  });
+  edges.push({ from: 'api', to: 'database', protocol: 'TCP' });
+
+  // Pattern-specific nodes
+  if (['event_driven', 'microservices', 'distributed_workers'].includes(patternKey)) {
+    nodes.push({
+      id: 'queue',
+      type: 'infrastructure',
+      technology: 'Message Queue',
+      responsibility: 'Async task routing',
+      scaling_model: 'partition-based',
+    });
+    nodes.push({
+      id: 'worker',
+      type: 'service',
+      technology: 'Background Worker',
+      responsibility: 'Async processing',
+      scaling_model: 'horizontal',
+    });
+    edges.push({ from: 'api', to: 'queue', protocol: 'AMQP/SQS' });
+    edges.push({ from: 'queue', to: 'worker', protocol: 'AMQP/SQS' });
+    edges.push({ from: 'worker', to: 'database', protocol: 'TCP' });
+  }
+
+  // Cache
+  if (
+    requirements &&
+    requirements.functional &&
+    requirements.functional.some(f => f.domain === 'performance' || f.req === 'caching layer')
+  ) {
+    nodes.push({
+      id: 'cache',
+      type: 'infrastructure',
+      technology: 'Redis',
+      responsibility: 'Query caching, session storage',
+      scaling_model: 'horizontal',
+    });
+    edges.push({ from: 'api', to: 'cache', protocol: 'TCP' });
+  }
+
+  // External services
+  if (
+    requirements &&
+    requirements.functional &&
+    requirements.functional.some(f => f.domain === 'financial')
+  ) {
+    nodes.push({
+      id: 'payment_gateway',
+      type: 'external',
+      technology: 'Stripe/Payment Provider',
+      responsibility: 'Payment processing',
+      scaling_model: 'managed',
+    });
+    edges.push({ from: 'api', to: 'payment_gateway', protocol: 'HTTPS' });
+  }
+
+  return { nodes, edges, pattern: candidate.name || patternKey };
+}
+
+// ============================================================================
+// 12. ARCHITECTURE DIFFERENCE ENGINE (SECTION 40)
+// ============================================================================
+
+/**
+ * Compare two architectures and identify material differences.
+ */
+function diffArchitectures(archA, archB) {
+  const diff = {
+    architectures: { a: archA.name || 'A', b: archB.name || 'B' },
+    components_added: [],
+    components_removed: [],
+    dependencies_changed: [],
+    failure_modes_changed: [],
+    consistency_changed: false,
+    cost_changed: false,
+    operational_burden_changed: false,
+    security_boundaries_changed: false,
+    scaling_behavior_changed: false,
+    summary: '',
+  };
+
+  const patternA = ARCHITECTURE_PATTERNS[archA.pattern] || {};
+  const patternB = ARCHITECTURE_PATTERNS[archB.pattern] || {};
+
+  // Compare required capabilities
+  const capsA = new Set(patternA.required_capabilities || []);
+  const capsB = new Set(patternB.required_capabilities || []);
+  diff.components_added = [...capsB].filter(c => !capsA.has(c));
+  diff.components_removed = [...capsA].filter(c => !capsB.has(c));
+
+  // Compare failure modes
+  const fmA = new Set(patternA.failure_modes || []);
+  const fmB = new Set(patternB.failure_modes || []);
+  diff.failure_modes_changed = [...fmB].filter(f => !fmA.has(f));
+
+  // Compare operational cost
+  diff.cost_changed = patternA.operational_cost !== patternB.operational_cost;
+  diff.operational_burden_changed = patternA.operational_cost !== patternB.operational_cost;
+
+  // Compare scaling
+  const scaleA = JSON.stringify(patternA.scaling_characteristics || {});
+  const scaleB = JSON.stringify(patternB.scaling_characteristics || {});
+  diff.scaling_behavior_changed = scaleA !== scaleB;
+
+  // Compare security
+  const secA = (patternA.security_considerations || []).length;
+  const secB = (patternB.security_considerations || []).length;
+  diff.security_boundaries_changed = secA !== secB;
+
+  diff.summary =
+    `Moving from ${diff.architectures.a} to ${diff.architectures.b}: ` +
+    `+${diff.components_added.length} components, -${diff.components_removed.length} components, ` +
+    `${diff.failure_modes_changed.length} new failure modes.`;
+
+  return diff;
+}
+
+// ============================================================================
+// 13. DECISION REVERSIBILITY & CONFIDENCE (SECTIONS 31–32)
+// ============================================================================
+
+/**
+ * Classify architecture decision reversibility.
+ */
+function classifyArchitectureReversibility(decisionType) {
+  const classifications = {
+    ui_library: 'EASILY_REVERSIBLE',
+    css_framework: 'EASILY_REVERSIBLE',
+    api_framework: 'MODERATELY_REVERSIBLE',
+    database_choice: 'HARD_TO_REVERSE',
+    data_model: 'HARD_TO_REVERSE',
+    consistency_model: 'HARD_TO_REVERSE',
+    api_contract: 'MODERATELY_REVERSIBLE',
+    deployment_platform: 'MODERATELY_REVERSIBLE',
+    architecture_pattern: 'HARD_TO_REVERSE',
+    message_broker: 'MODERATELY_REVERSIBLE',
+    auth_provider: 'MODERATELY_REVERSIBLE',
+    cloud_provider: 'HARD_TO_REVERSE',
+  };
+  return classifications[decisionType] || 'UNKNOWN';
+}
+
+/**
+ * Assess architecture confidence across dimensions.
+ */
+function assessArchitectureConfidence(requirements) {
+  const confidence = {};
+  const constraints = (requirements && requirements.constraints) || [];
+  const unknowns = (requirements && requirements.unknowns) || [];
+
+  confidence.database = constraints.some(c => c.category === 'technology') ? 'KNOWN' : 'ESTIMATED';
+  confidence.traffic = constraints.some(c => c.category === 'scale' || c.category === 'performance')
+    ? 'SUPPORTED'
+    : 'UNKNOWN';
+  confidence.team = constraints.some(c => c.category === 'team') ? 'KNOWN' : 'UNKNOWN';
+  confidence.budget = constraints.some(c => c.category === 'cost') ? 'KNOWN' : 'UNKNOWN';
+  confidence.compliance = constraints.some(c => c.category === 'governance') ? 'KNOWN' : 'UNKNOWN';
+  confidence.geographic = constraints.some(c => c.req === 'data residency') ? 'KNOWN' : 'UNKNOWN';
+
+  return confidence;
+}
+
+// ============================================================================
+// 14. FITNESS TEST GENERATOR (SECTION 42)
+// ============================================================================
+
+/**
+ * Generate architecture fitness tests for a candidate + ADR.
+ */
+function generateFitnessTests(candidate, adr) {
+  const tests = [];
+  const patternKey = candidate.pattern || '';
+
+  // Universal tests
+  tests.push({
+    name: 'no_circular_dependencies',
+    description: 'No circular dependency chains between modules/services',
+    type: 'structural',
+    automated: true,
+  });
+  tests.push({
+    name: 'no_forbidden_dependencies',
+    description: 'No module imports from forbidden layers (e.g., infrastructure importing domain)',
+    type: 'structural',
+    automated: true,
+  });
+  tests.push({
+    name: 'no_missing_timeouts',
+    description: 'All external HTTP/RPC calls have explicit timeout configuration',
+    type: 'resilience',
+    automated: true,
+  });
+  tests.push({
+    name: 'no_unbounded_retries',
+    description: 'All retry logic has maximum attempt limits and backoff',
+    type: 'resilience',
+    automated: true,
+  });
+
+  // Pattern-specific tests
+  if (['modular_monolith', 'hexagonal'].includes(patternKey)) {
+    tests.push({
+      name: 'module_boundary_enforcement',
+      description: 'Modules access each other only through defined interfaces, not direct imports',
+      type: 'structural',
+      automated: true,
+    });
+  }
+
+  if (['microservices', 'event_driven'].includes(patternKey)) {
+    tests.push({
+      name: 'no_shared_database',
+      description: 'Each service owns its database; no direct cross-service DB access',
+      type: 'data_ownership',
+      automated: true,
+    });
+    tests.push({
+      name: 'no_synchronous_chain',
+      description: 'No synchronous call chain exceeding 3 services deep',
+      type: 'resilience',
+      automated: true,
+    });
+  }
+
+  if (adr && adr.security_considerations && adr.security_considerations.length > 0) {
+    tests.push({
+      name: 'auth_before_business_logic',
+      description:
+        'All API endpoints enforce authentication and authorization before executing business logic',
+      type: 'security',
+      automated: true,
+    });
+  }
+
+  return tests;
+}
+
+// ============================================================================
+// 15. DRIFT DETECTION (SECTION 41)
+// ============================================================================
+
+/**
+ * Detect architectural drift between implementation and ADR.
+ */
+function detectArchitecturalDrift(implementation, adr) {
+  const drifts = [];
+
+  if (!implementation || !adr) return drifts;
+
+  // Check for unauthorized dependencies
+  const implDeps = implementation.dependencies || [];
+  const adrDeps = (adr.decision && adr.decision.rationale) || [];
+
+  // Check for new direct database access bypassing intended boundaries
+  if (implementation.direct_db_access_from_unauthorized_layer) {
+    drifts.push({
+      type: 'BOUNDARY_VIOLATION',
+      severity: 'HIGH',
+      description:
+        'Direct database access detected from unauthorized layer, bypassing repository boundary.',
+      adr_reference: adr.id || 'N/A',
+    });
+  }
+
+  // Check for new service dependencies creating synchronous cycles
+  if (implementation.synchronous_cycle_detected) {
+    drifts.push({
+      type: 'DEPENDENCY_CYCLE',
+      severity: 'CRITICAL',
+      description:
+        'New synchronous service dependency creates a cycle, contradicting ADR architecture.',
+      adr_reference: adr.id || 'N/A',
+    });
+  }
+
+  // Check for new cache creating second source of truth
+  if (implementation.new_cache_without_invalidation) {
+    drifts.push({
+      type: 'DATA_INTEGRITY',
+      severity: 'HIGH',
+      description:
+        'New cache introduced without invalidation strategy creates second source of truth.',
+      adr_reference: adr.id || 'N/A',
+    });
+  }
+
+  // Check for new API bypassing auth middleware
+  if (implementation.api_bypasses_auth) {
+    drifts.push({
+      type: 'SECURITY_VIOLATION',
+      severity: 'CRITICAL',
+      description: 'New API endpoint bypasses authentication/authorization middleware.',
+      adr_reference: adr.id || 'N/A',
+    });
+  }
+
+  return drifts;
+}
+
+// ============================================================================
+// 16. PHASE 7 METRICS TRACKER (SECTION 45)
+// ============================================================================
+
+class Phase7MetricsTracker {
+  constructor() {
+    this.metrics = {
+      requirement_extraction_accuracy: { correct: 0, total: 0 },
+      constraint_discovery_rate: { discovered: 0, total: 0 },
+      candidate_relevance: { relevant: 0, total: 0 },
+      tradeoff_coverage: { covered: 0, total: 0 },
+      unsupported_assumption_detection: { detected: 0, total: 0 },
+      architecture_failure_discovery: { discovered: 0, total: 0 },
+      adr_completeness: { complete_fields: 0, total_fields: 0 },
+      drift_detection: { detected: 0, total: 0 },
+      decision_reversal_detection: { detected: 0, total: 0 },
+      simulation_decision_coverage: { covered: 0, total: 0 },
+      historical_reuse_accuracy: { accurate: 0, total: 0 },
+    };
+  }
+
+  record(metric, success) {
+    if (this.metrics[metric]) {
+      this.metrics[metric].total++;
+      if (success) {
+        const key = Object.keys(this.metrics[metric]).find(k => k !== 'total');
+        if (key) this.metrics[metric][key]++;
+      }
+    }
+  }
+
+  getMetrics() {
+    const result = {};
+    for (const [key, val] of Object.entries(this.metrics)) {
+      result[key] = {
+        ...val,
+        rate: val.total > 0 ? val[Object.keys(val).find(k => k !== 'total')] / val.total : 0,
+      };
+    }
+    return result;
+  }
+}
+
+let _phase7MetricsTracker = null;
+
+function getPhase7MetricsTracker() {
+  if (!_phase7MetricsTracker) {
+    _phase7MetricsTracker = new Phase7MetricsTracker();
+  }
+  return _phase7MetricsTracker;
+}
+
+// ============================================================================
+// 17. PROGRESSIVE ANALYSIS (SECTION 46)
+// ============================================================================
+
+/**
+ * Determine how deep the architecture analysis should go.
+ *
+ * @param {string} taskQuery
+ * @returns {string} 'NONE' | 'MINIMAL' | 'STANDARD' | 'FULL'
+ */
+function determineAnalysisDepth(taskQuery) {
+  const q = (taskQuery || '').toLowerCase();
+
+  // Tier 0: No architecture analysis needed
+  if (/(?:^fix|^typo|^css|^format|^rename|^lint|^style|^readme|^comment|^doc)/i.test(q.trim())) {
+    return 'NONE';
+  }
+
+  // Tier 1: Minimal — single file change, simple logic
+  if (/(?:^add.*button|^change.*color|^update.*text|^modify.*config|^add.*field)/i.test(q.trim())) {
+    return 'MINIMAL';
+  }
+
+  // Tier 3: Full — architectural keywords present
+  if (
+    /(?:architect|design.*system|multi.?tenant|distributed|microservice|platform|saas|scale|million|enterprise|migrate)/i.test(
+      q,
+    )
+  ) {
+    return 'FULL';
+  }
+
+  // Tier 2: Standard — multi-component work
+  if (/(?:build|create|implement|api|backend|frontend|database|service|application|app)/i.test(q)) {
+    return 'STANDARD';
+  }
+
+  return 'MINIMAL';
+}
+
+// ============================================================================
+// 18. MAIN ORCHESTRATOR — runArchitectureAnalysis
+// ============================================================================
+
+/**
+ * Run the full Phase 7 architecture intelligence pipeline.
+ *
+ * @param {string} taskQuery
+ * @param {Object} [options]
+ * @returns {Object} Full architecture analysis
+ */
+function runArchitectureAnalysis(taskQuery, options = {}) {
+  const depth = options.depth || determineAnalysisDepth(taskQuery);
+
+  if (depth === 'NONE') {
+    return {
+      task: taskQuery,
+      depth: 'NONE',
+      skipped: true,
+      reason: 'Task does not require architecture analysis (Tier 0).',
+      requirements: null,
+      candidates: null,
+      tradeoffs: null,
+      adr: null,
+    };
+  }
+
+  // Step 1: Extract requirements
+  const requirements = extractRequirements(taskQuery);
+
+  // Step 2: Extract constraints
+  const constraintResult = extractConstraints(taskQuery, requirements);
+
+  if (depth === 'MINIMAL') {
+    return {
+      task: taskQuery,
+      depth: 'MINIMAL',
+      requirements,
+      constraints: constraintResult,
+      candidates: null,
+      tradeoffs: null,
+      adr: null,
+      anti_patterns: [],
+      reason: 'Task requires minimal architecture analysis (Tier 1).',
+    };
+  }
+
+  // Step 3: Detect anti-patterns
+  const antiPatterns = detectAntiPatterns(taskQuery, requirements);
+
+  // Step 4: Generate candidates
+  const candidateResult = generateCandidates(requirements, constraintResult);
+
+  // Step 5: Build trade-offs
+  let tradeoffResult = null;
+  if (candidateResult.candidates.length >= 2) {
+    tradeoffResult = analyzeTradeoffs(candidateResult.candidates, requirements, constraintResult);
+  }
+
+  // Step 6: Decision sensitivity
+  const sensitivity = analyzeDecisionSensitivity(
+    candidateResult.candidates,
+    tradeoffResult,
+    requirements,
+  );
+
+  // Step 7: Breakpoints for top candidate
+  const breakpoints = candidateResult.candidates[0]
+    ? identifyBreakpoints(candidateResult.candidates[0], requirements)
+    : [];
+
+  // Step 8: Domain-specific analysis
+  const buildVsBuy = analyzeBuildVsBuy(taskQuery, requirements);
+  const dataArchitecture = analyzeDataArchitecture(taskQuery, requirements);
+  const communicationPatterns = analyzeCommunicationPatterns(taskQuery);
+  const securityArchitecture = analyzeSecurityArchitecture(taskQuery, requirements);
+
+  // Step 9: Cost & operational analysis for top candidate
+  let costAnalysis = null;
+  let operationalAnalysis = null;
+  if (candidateResult.candidates[0]) {
+    costAnalysis = analyzeCostArchitecture(candidateResult.candidates[0], constraintResult);
+    operationalAnalysis = analyzeOperationalComplexity(candidateResult.candidates[0]);
+  }
+
+  // Step 10: Evolution analysis
+  const evolutionAnalysis = candidateResult.candidates[0]
+    ? analyzeEvolution(candidateResult.candidates[0])
+    : null;
+
+  // Step 11: Architecture graph
+  const architectureGraph = candidateResult.candidates[0]
+    ? generateArchitectureGraph(candidateResult.candidates[0], requirements, securityArchitecture)
+    : null;
+
+  // Step 12: Confidence assessment
+  const confidence = assessArchitectureConfidence(requirements);
+
+  // Step 13: Fitness tests
+  let fitnessTests = [];
+  let adr = null;
+
+  if (depth === 'FULL' || candidateResult.candidates.length > 0) {
+    // Step 14: Generate ADR
+    adr = generateADR({
+      task: taskQuery,
+      requirements,
+      constraints: constraintResult,
+      candidates: candidateResult.candidates,
+      tradeoffs: tradeoffResult,
+      decision: candidateResult.candidates[0],
+      antiPatterns,
+      sensitivity,
+      breakpoints,
+      securityAnalysis: securityArchitecture,
+    });
+
+    fitnessTests = candidateResult.candidates[0]
+      ? generateFitnessTests(candidateResult.candidates[0], adr)
+      : [];
+  }
+
+  return {
+    task: taskQuery,
+    depth,
+    requirements,
+    constraints: constraintResult,
+    anti_patterns: antiPatterns,
+    candidates: candidateResult,
+    tradeoffs: tradeoffResult,
+    decision_sensitivity: sensitivity,
+    breakpoints,
+    build_vs_buy: buildVsBuy,
+    data_architecture: dataArchitecture,
+    communication_patterns: communicationPatterns,
+    security_architecture: securityArchitecture,
+    cost_analysis: costAnalysis,
+    operational_analysis: operationalAnalysis,
+    evolution: evolutionAnalysis,
+    architecture_graph: architectureGraph,
+    confidence,
+    fitness_tests: fitnessTests,
+    adr,
+    memory_updates: [],
+  };
+}
+
+// ============================================================================
+// EXPORTS
+// ============================================================================
+
+module.exports = {
+  // Constants
+  REQUIREMENT_TYPES,
+  CONFIDENCE_LEVELS,
+  ARCHITECTURE_PATTERNS,
+  ANTI_PATTERNS,
+  TRADEOFF_CRITERIA,
+
+  // Requirement Intelligence (§4–§5)
+  extractRequirements,
+
+  // Constraint Extraction (§6)
+  extractConstraints,
+
+  // Pattern Registry (§8)
+  getArchitecturePatterns,
+
+  // Anti-Pattern Detection (§9)
+  detectAntiPatterns,
+
+  // Candidate Generation (§7)
+  generateCandidates,
+
+  // Trade-Off Engine (§10–§11)
+  analyzeTradeoffs,
+  rateCriterion,
+  determineCriterionImportance,
+
+  // Decision Sensitivity & Breakpoints (§12–§13)
+  analyzeDecisionSensitivity,
+  identifyBreakpoints,
+
+  // Domain-Specific Analyzers (§14–§27)
+  analyzeBuildVsBuy,
+  analyzeDataArchitecture,
+  analyzeCommunicationPatterns,
+  analyzeSecurityArchitecture,
+  analyzeCostArchitecture,
+  analyzeOperationalComplexity,
+
+  // Migration & Evolution (§28–§29)
+  analyzeMigration,
+  analyzeEvolution,
+
+  // ADR Generation (§30)
+  generateADR,
+
+  // Architecture Graph (§39)
+  generateArchitectureGraph,
+
+  // Difference Engine (§40)
+  diffArchitectures,
+
+  // Reversibility & Confidence (§31–§32)
+  classifyArchitectureReversibility,
+  assessArchitectureConfidence,
+
+  // Fitness Tests (§42)
+  generateFitnessTests,
+
+  // Drift Detection (§41)
+  detectArchitecturalDrift,
+
+  // Metrics (§45)
+  Phase7MetricsTracker,
+  getPhase7MetricsTracker,
+
+  // Progressive Analysis (§46)
+  determineAnalysisDepth,
+
+  // Main Orchestrator
+  runArchitectureAnalysis,
+};