agentshield-sdk 7.0.0

package/src/main.js

@@ -0,0 +1,741 @@
+'use strict';
+
+/**
+ * Agent Shield — Unified Entry Point
+ *
+ * Import everything from a single module:
+ *   const shield = require('agent-shield');
+ *
+ * Each module is loaded safely — if one fails, the rest still work.
+ */
+
+/**
+ * Safely require a module. Returns its exports or an empty object on failure.
+ * @param {string} path - Module path
+ * @param {string} label - Label for error logging
+ * @returns {object}
+ */
+function safeRequire(path, label) {
+  try {
+    return require(path);
+  } catch (err) {
+    console.warn(`[Agent Shield] Failed to load ${label}: ${err.message}`);
+    return {};
+  }
+}
+
+// Core (these are critical — if they fail, we still export what we can)
+const { AgentShield } = safeRequire('./index', 'core');
+const { scanText, getPatterns, SEVERITY_ORDER } = safeRequire('./detector-core', 'detector-core');
+const { expressMiddleware, wrapAgent, shieldTools, extractTextFromBody } = safeRequire('./middleware', 'middleware');
+
+// Protection
+const { CircuitBreaker, shadowMode, RateLimiter, STATE } = safeRequire('./circuit-breaker', 'circuit-breaker');
+const { CanaryTokens, PromptLeakDetector, API_KEY_PATTERNS } = safeRequire('./canary', 'canary');
+const { PIIRedactor, DLPEngine, ContentPolicy, PII_PATTERNS, CONTENT_CATEGORIES } = safeRequire('./pii', 'pii');
+const { ToolSequenceAnalyzer, PermissionBoundary, InputQuarantine, SUSPICIOUS_SEQUENCES } = safeRequire('./tool-guard', 'tool-guard');
+
+// Conversation
+const { FragmentationDetector, LanguageSwitchDetector, TokenBudgetAnalyzer, InstructionHierarchy, BehavioralFingerprint } = safeRequire('./conversation', 'conversation');
+
+// Policy & Logging
+const { loadPolicy, loadPolicyFile, StructuredLogger, WebhookAlert, LOG_LEVEL } = safeRequire('./policy', 'policy');
+
+// Multi-Agent
+const { AgentFirewall, DelegationChain, SharedThreatState } = safeRequire('./multi-agent', 'multi-agent');
+
+// Advanced Detection
+const { SteganographyDetector, EncodingBruteforceDetector, StructuredDataScanner, STEGO_PATTERNS } = safeRequire('./encoding', 'encoding');
+
+// Watermarking & Privacy
+const { OutputWatermark, DifferentialPrivacy } = safeRequire('./watermark', 'watermark');
+
+// Utilities
+const { getGrade, getGradeLabel, makeBar, truncate, formatHeader, generateId } = safeRequire('./utils', 'utils');
+
+// Error codes & deprecation
+const { ERROR_CODES, createShieldError, deprecationWarning } = safeRequire('./errors', 'errors');
+
+// v7.0 — MCP Security Runtime
+const { MCPSecurityRuntime, MCPSessionStateMachine, SESSION_STATES } = safeRequire('./mcp-security-runtime', 'mcp-security-runtime');
+
+// v7.0 — MCP SDK Integration
+const { shieldMCPServer, createMCPSecurityLayer } = safeRequire('./mcp-sdk-integration', 'mcp-sdk-integration');
+
+// v7.0 — MCP Certification & Trust
+const { AgentThreatIntelligence, MCPCertification, CrossOrgAgentTrust, THREAT_CATEGORIES: CERT_THREAT_CATEGORIES, CERTIFICATION_REQUIREMENTS, CERTIFICATION_LEVELS } = safeRequire('./mcp-certification', 'mcp-certification');
+
+// v5.1 — Stream scanning
+const { StreamScanner, createStreamWrapper, scanAsyncIterator, StreamBuffer } = safeRequire('./stream-scanner', 'stream-scanner');
+
+// v5.1 — Immutable audit log
+const { ImmutableAuditLog, AuditEntry, MemoryAuditStore, FileAuditStore, AuditProof, verifyAuditChain } = safeRequire('./audit-immutable', 'audit-immutable');
+
+// v5.1 — Agent observability
+const { PrometheusExporter, DatadogLogger, MetricsCollector: ObservabilityMetrics } = safeRequire('./observability', 'observability');
+
+// v5.1 — Benchmark harness
+const { BenchmarkHarness, DatasetLoader, BenchmarkMetrics, RegressionTracker, BenchmarkReportGenerator } = safeRequire('./benchmark-harness', 'benchmark-harness');
+
+// Integrations
+const { ShieldCallbackHandler, shieldAnthropicClient, shieldOpenAIClient, shieldVercelAI, shieldFetch, ShieldBlockError } = safeRequire('./integrations', 'integrations');
+
+// Red Team
+const { AttackSimulator, PayloadFuzzer, getAttackCategories, getPayloads, ATTACK_PAYLOADS } = safeRequire('./redteam', 'redteam');
+
+// Shield Score
+const { ShieldScoreCalculator, SCORE_CATEGORIES } = safeRequire('./shield-score', 'shield-score');
+
+// Benchmark Harness — imported above (line 69)
+
+// Threat Encyclopedia
+const { ThreatEncyclopedia, THREAT_ENCYCLOPEDIA, DAILY_PATTERNS } = safeRequire('./threat-encyclopedia', 'threat-encyclopedia');
+
+// Compliance & Audit
+const { ComplianceReporter, AuditTrail, IncidentPlaybook, SecurityChecklistGenerator, COMPLIANCE_FRAMEWORKS, INCIDENT_PLAYBOOKS } = safeRequire('./compliance', 'compliance');
+
+// Enterprise
+const { MultiTenantShield, RoleBasedPolicy, DebugShield, DEFAULT_ROLES } = safeRequire('./enterprise', 'enterprise');
+
+// Badges
+const { BadgeGenerator, GitHubActionReporter } = safeRequire('./badges', 'badges');
+
+// Allowlist & Feedback
+const { Allowlist, ConfidenceCalibrator, FeedbackLoop, ScanCache } = safeRequire('./allowlist', 'allowlist');
+
+// Presets & Config Builder
+const { PRESETS, ConfigBuilder, SnippetGenerator, getPresets, getPreset } = safeRequire('./presets', 'presets');
+
+// Advanced Scanners
+const { RAGScanner, RAG_INJECTION_PATTERNS, PromptLinter, LINT_RULES, ToolSchemaValidator, DANGEROUS_TOOL_PATTERNS } = safeRequire('./scanners', 'scanners');
+
+// Production
+const { SamplingScanner, ShadowComparison, GracefulScanner, ThreatReplay, AttackAttributionChain, DiffReporter, PostureTracker } = safeRequire('./production', 'production');
+
+// Testing & Contracts
+const { TestSuiteGenerator, ATTACK_TEMPLATES, AgentContract, BreakglassProtocol } = safeRequire('./testing', 'testing');
+
+// Multi-Agent Trust
+const { MessageSigner, CapabilityToken, DelegationManager, BlastRadiusContainer } = safeRequire('./multi-agent-trust', 'multi-agent-trust');
+
+// Extended Policy & Intelligence
+const { ABTestRunner, ThreatIntelFeed, PatternBuilder, Doctor, GitHubActionGenerator, SOCIntegration, MigrationGuide, Playground } = safeRequire('./policy-extended', 'policy-extended');
+
+// --- New Modules ---
+
+// Streaming (legacy import — full import on line 60)
+// StreamScanner, StreamBuffer, createStreamWrapper, scanAsyncIterator loaded above
+
+// Plugin System
+const { PluginManager, PluginTemplate, PluginSandbox } = safeRequire('./plugin-system', 'plugin-system');
+
+// Token Analysis
+const { EntropyAnalyzer, PerplexityEstimator, BurstDetector, TextStatistics } = safeRequire('./token-analysis', 'token-analysis');
+
+// Document Scanner
+const { DocumentScanner, TextExtractor, IndirectInjectionScanner } = safeRequire('./document-scanner', 'document-scanner');
+
+// Tool Output Validator
+const { ToolOutputValidator, OutputSanitizer } = safeRequire('./tool-output-validator', 'tool-output-validator');
+
+// Response Handler
+const { ResponseHandler, ResponseTemplates, ReviewQueue } = safeRequire('./response-handler', 'response-handler');
+
+// Worker Scanner
+const { WorkerScanner, ScanQueue, ThreadedWorkerScanner } = safeRequire('./worker-scanner', 'worker-scanner');
+
+// Alert Tuning
+const { AlertFatigueAnalyzer, AutoTuner, AlertCorrelator } = safeRequire('./alert-tuning', 'alert-tuning');
+
+// OpenTelemetry
+const { ShieldMetrics, ShieldTracer, MetricsDashboard } = safeRequire('./otel', 'otel');
+
+// Certification
+const { CertificationRunner, Certificate, CertificationHistory } = safeRequire('./certification', 'certification');
+
+// MCP Server
+const { MCPServer, MCPToolHandler } = safeRequire('./mcp-server', 'mcp-server');
+
+// CTF
+const { CTFEngine, CTFReporter, CHALLENGES } = safeRequire('./ctf', 'ctf');
+
+// Observability — imported above (line 66)
+
+// Adaptive Detection
+const { AdaptiveDetector, SemanticAnalysisHook, CommunityPatterns } = safeRequire('./adaptive', 'adaptive');
+
+// OpenClaw
+const { OpenClawShieldSkill, shieldOpenClawMessages, generateOpenClawSkill } = safeRequire('./openclaw', 'openclaw');
+
+// --- v1.2 Modules ---
+
+// Semantic Detection
+const { SemanticClassifier, httpPost } = safeRequire('./semantic', 'semantic');
+
+// Embedding Similarity
+const { EmbeddingSimilarityDetector, ATTACK_CORPUS, tokenize: tokenizeText, cosineSimilarity } = safeRequire('./embedding', 'embedding');
+
+// Context-Aware Scoring
+const { ConversationContextAnalyzer, ESCALATION_SIGNALS, TOPIC_PIVOT_SIGNALS } = safeRequire('./context-scoring', 'context-scoring');
+
+// Confidence Tuning
+const { ConfidenceTuner } = safeRequire('./confidence-tuning', 'confidence-tuning');
+
+// --- v2.0 Modules ---
+
+// Plugin Marketplace
+const { PluginRegistry, PluginValidator, MarketplaceClient } = safeRequire('./plugin-marketplace', 'plugin-marketplace');
+
+// --- v2.1 Modules ---
+
+// Distributed Scanning
+const { DistributedShield, DistributedAdapter, MemoryAdapter, RedisAdapter } = safeRequire('./distributed', 'distributed');
+
+// Audit Log Streaming
+const { AuditStreamManager, AuditTransport, FileTransport, SplunkTransport, ElasticsearchTransport } = safeRequire('./audit-streaming', 'audit-streaming');
+
+// Immutable Audit Log — imported above (line 63)
+
+// --- v3.0 Modules ---
+
+// Self-Healing Patterns
+const { SelfHealingEngine, PatternGenerator: SelfHealingPatternGenerator } = safeRequire('./self-healing', 'self-healing');
+
+// Honeypot Mode
+const { HoneypotEngine, HoneypotSession } = safeRequire('./honeypot', 'honeypot');
+
+// Multi-Modal Scanning
+const { MultiModalScanner, ModalityExtractor } = safeRequire('./multimodal', 'multimodal');
+
+// Behavior Profiling
+const { BehaviorProfile } = safeRequire('./behavior-profiling', 'behavior-profiling');
+
+// --- Remaining Roadmap Modules ---
+
+// SSO/SAML Integration
+const { SSOManager, SAMLParser, OIDCHandler, IdentityMapper, SSOSession, DEFAULT_MAPPINGS: SSO_DEFAULT_MAPPINGS } = safeRequire('./sso-saml', 'sso-saml');
+
+// Custom Model Fine-Tuning
+const { ModelTrainer, TrainingPipeline, DatasetManager, ModelEvaluator, FineTunedModel } = safeRequire('./model-finetuning', 'model-finetuning');
+
+// Threat Intelligence Network
+const { ThreatIntelNetwork, PeerNode, PatternAnonymizer, ConsensusEngine, ThreatFeed, NETWORK_DEFAULTS } = safeRequire('./threat-intel-network', 'threat-intel-network');
+
+// --- v4.0 Modules ---
+
+// Multi-Language Patterns
+const { I18nPatternManager, CJK_PATTERNS, ARABIC_PATTERNS, CYRILLIC_PATTERNS, INDIC_PATTERNS, MULTILINGUAL_PATTERNS, getI18nPatterns } = safeRequire('./i18n-patterns', 'i18n-patterns');
+
+// LLM Red Team Suite
+const { LLMRedTeamSuite, AdversarialGenerator, JailbreakLibrary, EvasionTester, RedTeamReport, JAILBREAK_TEMPLATES, MUTATION_TECHNIQUES } = safeRequire('./llm-redteam', 'llm-redteam');
+
+// --- v5.0 Modules ---
+
+// Agent-to-Agent Protocol
+const { AgentProtocol, SecureChannel, HandshakeManager, AgentIdentity, ProtocolMessage, MessageRouter, PROTOCOL_VERSION } = safeRequire('./agent-protocol', 'agent-protocol');
+
+// Policy-as-Code DSL
+const { PolicyDSL, PolicyParser, PolicyCompiler, PolicyRuntime, PolicyValidator, BUILTIN_FUNCTIONS: DSL_BUILTINS, EXAMPLE_STRICT_POLICY, EXAMPLE_PERMISSIVE_POLICY, EXAMPLE_CUSTOM_RULES_POLICY } = safeRequire('./policy-dsl', 'policy-dsl');
+
+// Fuzzing Harness
+const { FuzzingHarness, InputGenerator, MutationEngine: FuzzMutationEngine, CoverageTracker, FuzzReport, CrashCollector, SEED_CORPUS } = safeRequire('./fuzzer', 'fuzzer');
+
+// Model Fingerprinting
+const { ModelFingerprinter, ResponseAnalyzer, StyleProfile, FingerprintDatabase, SupplyChainDetector, MODEL_SIGNATURES } = safeRequire('./model-fingerprint', 'model-fingerprint');
+
+// Cost/Latency Optimizer
+const { CostOptimizer, LatencyBudget, AdaptiveScanner, TierManager, PerformanceMonitor, ScanPlan, OPTIMIZATION_PRESETS } = safeRequire('./cost-optimizer', 'cost-optimizer');
+
+// --- v6.0 — Compliance & Market Readiness ---
+
+// OWASP LLM Top 10 v2025 Coverage Matrix
+const { OWASP_LLM_2025, OWASPCoverageMatrix, SEVERITY_WEIGHTS: OWASP_SEVERITY_WEIGHTS, COVERAGE_MULTIPLIERS } = safeRequire('./owasp-2025', 'owasp-2025');
+
+// MCP Bridge (Model Context Protocol integration)
+const { MCPBridge, MCPToolPolicy, MCPSessionGuard, MCPResourceScanner, MCP_DANGEROUS_TOOLS, ARG_INJECTION_PATTERNS, createMCPMiddleware } = safeRequire('./mcp-bridge', 'mcp-bridge');
+
+// NIST AI RMF Mapping & AI-BOM Generator
+const { NIST_AI_RMF_2025, SP800_53_AI_CONTROLS, NISTMapper, AIBOMGenerator, ComplianceChecker: NISTComplianceChecker } = safeRequire('./nist-mapping', 'nist-mapping');
+
+// EU AI Act Compliance
+const { EU_AI_ACT_REQUIREMENTS, RiskClassifier, ConformityAssessment, TransparencyReporter, IncidentReporter: EUIncidentReporter, EUAIActDashboard } = safeRequire('./eu-ai-act', 'eu-ai-act');
+
+// System Prompt Leakage Detector (OWASP LLM07-2025)
+const { PROMPT_EXTRACTION_PATTERNS, SystemPromptGuard, PromptFingerprinter, PromptLeakageMitigation } = safeRequire('./prompt-leakage', 'prompt-leakage');
+
+// RAG/Vector Vulnerability Scanner (OWASP LLM08-2025)
+const { RAG_VULNERABILITY_PATTERNS, VECTOR_DB_SECURITY_CHECKLIST, RAGVulnerabilityScanner, EmbeddingIntegrityChecker, RAGPipelineAuditor } = safeRequire('./rag-vulnerability', 'rag-vulnerability');
+
+// Confused Deputy Prevention (Meta Incident Response)
+const { AuthorizationContext, EphemeralTokenManager, IntentValidator, ConfusedDeputyGuard } = safeRequire('./confused-deputy', 'confused-deputy');
+
+// Build exports, filtering out undefined values from failed imports
+const _exports = {
+  // Core
+  AgentShield,
+  scanText,
+  getPatterns,
+  SEVERITY_ORDER,
+
+  // Middleware
+  expressMiddleware,
+  wrapAgent,
+  shieldTools,
+  extractTextFromBody,
+
+  // Protection
+  CircuitBreaker,
+  shadowMode,
+  RateLimiter,
+  STATE,
+  CanaryTokens,
+  PromptLeakDetector,
+  API_KEY_PATTERNS,
+  PIIRedactor,
+  DLPEngine,
+  ContentPolicy,
+  PII_PATTERNS,
+  CONTENT_CATEGORIES,
+  ToolSequenceAnalyzer,
+  PermissionBoundary,
+  InputQuarantine,
+  SUSPICIOUS_SEQUENCES,
+
+  // Conversation
+  FragmentationDetector,
+  LanguageSwitchDetector,
+  TokenBudgetAnalyzer,
+  InstructionHierarchy,
+  BehavioralFingerprint,
+
+  // Policy & Logging
+  loadPolicy,
+  loadPolicyFile,
+  StructuredLogger,
+  WebhookAlert,
+  LOG_LEVEL,
+
+  // Multi-Agent
+  AgentFirewall,
+  DelegationChain,
+  SharedThreatState,
+
+  // Advanced Detection
+  SteganographyDetector,
+  EncodingBruteforceDetector,
+  StructuredDataScanner,
+  STEGO_PATTERNS,
+
+  // Watermarking & Privacy
+  OutputWatermark,
+  DifferentialPrivacy,
+
+  // Utilities
+  getGrade,
+  getGradeLabel,
+  makeBar,
+  truncate,
+  formatHeader,
+  generateId,
+
+  // Integrations
+  ShieldCallbackHandler,
+  shieldAnthropicClient,
+  shieldOpenAIClient,
+  shieldVercelAI,
+  shieldFetch,
+  ShieldBlockError,
+
+  // Red Team
+  AttackSimulator,
+  PayloadFuzzer,
+  getAttackCategories,
+  getPayloads,
+  ATTACK_PAYLOADS,
+
+  // Shield Score
+  ShieldScoreCalculator,
+  SCORE_CATEGORIES,
+
+  // Benchmark Harness
+  BenchmarkHarness,
+  DatasetLoader,
+  BenchmarkMetrics,
+  RegressionTracker,
+  BenchmarkReportGenerator,
+
+  // Threat Encyclopedia
+  ThreatEncyclopedia,
+  THREAT_ENCYCLOPEDIA,
+  DAILY_PATTERNS,
+
+  // Compliance & Audit
+  ComplianceReporter,
+  AuditTrail,
+  IncidentPlaybook,
+  SecurityChecklistGenerator,
+  COMPLIANCE_FRAMEWORKS,
+  INCIDENT_PLAYBOOKS,
+
+  // Enterprise
+  MultiTenantShield,
+  RoleBasedPolicy,
+  DebugShield,
+  DEFAULT_ROLES,
+
+  // Badges
+  BadgeGenerator,
+  GitHubActionReporter,
+
+  // Allowlist & Feedback
+  Allowlist,
+  ConfidenceCalibrator,
+  FeedbackLoop,
+  ScanCache,
+
+  // Presets & Config Builder
+  PRESETS,
+  ConfigBuilder,
+  SnippetGenerator,
+  getPresets,
+  getPreset,
+
+  // Advanced Scanners
+  RAGScanner,
+  RAG_INJECTION_PATTERNS,
+  PromptLinter,
+  LINT_RULES,
+  ToolSchemaValidator,
+  DANGEROUS_TOOL_PATTERNS,
+
+  // Production
+  SamplingScanner,
+  ShadowComparison,
+  GracefulScanner,
+  ThreatReplay,
+  AttackAttributionChain,
+  DiffReporter,
+  PostureTracker,
+
+  // Testing & Contracts
+  TestSuiteGenerator,
+  ATTACK_TEMPLATES,
+  AgentContract,
+  BreakglassProtocol,
+
+  // Multi-Agent Trust
+  MessageSigner,
+  CapabilityToken,
+  DelegationManager,
+  BlastRadiusContainer,
+
+  // Extended Policy & Intelligence
+  ABTestRunner,
+  ThreatIntelFeed,
+  PatternBuilder,
+  Doctor,
+  GitHubActionGenerator,
+  SOCIntegration,
+  MigrationGuide,
+  Playground,
+
+  // Streaming
+  StreamScanner,
+  StreamBuffer,
+  createStreamWrapper,
+  scanAsyncIterator,
+
+  // Plugin System
+  PluginManager,
+  PluginTemplate,
+  PluginSandbox,
+
+  // Token Analysis
+  EntropyAnalyzer,
+  PerplexityEstimator,
+  BurstDetector,
+  TextStatistics,
+
+  // Document Scanner
+  DocumentScanner,
+  TextExtractor,
+  IndirectInjectionScanner,
+
+  // Tool Output Validator
+  ToolOutputValidator,
+  OutputSanitizer,
+
+  // Response Handler
+  ResponseHandler,
+  ResponseTemplates,
+  ReviewQueue,
+
+  // Worker Scanner
+  WorkerScanner,
+  ScanQueue,
+  ThreadedWorkerScanner,
+
+  // Alert Tuning
+  AlertFatigueAnalyzer,
+  AutoTuner,
+  AlertCorrelator,
+
+  // OpenTelemetry
+  ShieldMetrics,
+  ShieldTracer,
+  MetricsDashboard,
+
+  // Certification
+  CertificationRunner,
+  Certificate,
+  CertificationHistory,
+
+  // MCP Server
+  MCPServer,
+  MCPToolHandler,
+
+  // CTF
+  CTFEngine,
+  CTFReporter,
+  CHALLENGES,
+
+  // Observability
+  PrometheusExporter,
+  DatadogLogger,
+  MetricsCollector: ObservabilityMetrics,
+
+  // Adaptive Detection
+  AdaptiveDetector,
+  SemanticAnalysisHook,
+  CommunityPatterns,
+
+  // OpenClaw
+  OpenClawShieldSkill,
+  shieldOpenClawMessages,
+  generateOpenClawSkill,
+
+  // v1.2 — Semantic Detection
+  SemanticClassifier,
+  httpPost,
+  EmbeddingSimilarityDetector,
+  ATTACK_CORPUS,
+  tokenizeText,
+  cosineSimilarity,
+  ConversationContextAnalyzer,
+  ESCALATION_SIGNALS,
+  TOPIC_PIVOT_SIGNALS,
+  ConfidenceTuner,
+
+  // v2.0 — Plugin Marketplace
+  PluginRegistry,
+  PluginValidator,
+  MarketplaceClient,
+
+  // v2.1 — Distributed Scanning
+  DistributedShield,
+  DistributedAdapter,
+  MemoryAdapter,
+  RedisAdapter,
+
+  // v2.1 — Audit Log Streaming
+  AuditStreamManager,
+  AuditTransport,
+  FileTransport,
+  SplunkTransport,
+  ElasticsearchTransport,
+
+  // Immutable Audit Log — exported below in v5.1 section
+
+  // v3.0 — Self-Healing Patterns
+  SelfHealingEngine,
+  SelfHealingPatternGenerator,
+
+  // v3.0 — Honeypot Mode
+  HoneypotEngine,
+  HoneypotSession,
+
+  // v3.0 — Multi-Modal Scanning
+  MultiModalScanner,
+  ModalityExtractor,
+
+  // v3.0 — Behavior Profiling
+  BehaviorProfile,
+
+  // SSO/SAML Integration
+  SSOManager,
+  SAMLParser,
+  OIDCHandler,
+  IdentityMapper,
+  SSOSession,
+  SSO_DEFAULT_MAPPINGS,
+
+  // Custom Model Fine-Tuning
+  ModelTrainer,
+  TrainingPipeline,
+  DatasetManager,
+  ModelEvaluator,
+  FineTunedModel,
+
+  // Threat Intelligence Network
+  ThreatIntelNetwork,
+  PeerNode,
+  PatternAnonymizer,
+  ConsensusEngine,
+  ThreatFeed,
+  NETWORK_DEFAULTS,
+
+  // v4.0 — Multi-Language Patterns
+  I18nPatternManager,
+  CJK_PATTERNS,
+  ARABIC_PATTERNS,
+  CYRILLIC_PATTERNS,
+  INDIC_PATTERNS,
+  MULTILINGUAL_PATTERNS,
+  getI18nPatterns,
+
+  // v4.0 — LLM Red Team Suite
+  LLMRedTeamSuite,
+  AdversarialGenerator,
+  JailbreakLibrary,
+  EvasionTester,
+  RedTeamReport,
+  JAILBREAK_TEMPLATES,
+  MUTATION_TECHNIQUES,
+
+  // v5.0 — Agent-to-Agent Protocol
+  AgentProtocol,
+  SecureChannel,
+  HandshakeManager,
+  AgentIdentity,
+  ProtocolMessage,
+  MessageRouter,
+  PROTOCOL_VERSION,
+
+  // v5.0 — Policy-as-Code DSL
+  PolicyDSL,
+  PolicyParser,
+  PolicyCompiler,
+  PolicyRuntime,
+  PolicyValidator,
+  DSL_BUILTINS,
+  EXAMPLE_STRICT_POLICY,
+  EXAMPLE_PERMISSIVE_POLICY,
+  EXAMPLE_CUSTOM_RULES_POLICY,
+
+  // v5.0 — Fuzzing Harness
+  FuzzingHarness,
+  InputGenerator,
+  FuzzMutationEngine,
+  CoverageTracker,
+  FuzzReport,
+  CrashCollector,
+  SEED_CORPUS,
+
+  // v5.0 — Model Fingerprinting
+  ModelFingerprinter,
+  ResponseAnalyzer,
+  StyleProfile,
+  FingerprintDatabase,
+  SupplyChainDetector,
+  MODEL_SIGNATURES,
+
+  // v5.0 — Cost/Latency Optimizer
+  CostOptimizer,
+  LatencyBudget,
+  AdaptiveScanner,
+  TierManager,
+  PerformanceMonitor,
+  ScanPlan,
+  OPTIMIZATION_PRESETS,
+
+  // Error codes & deprecation
+  ERROR_CODES,
+  createShieldError,
+  deprecationWarning,
+
+  // v5.1 — Immutable audit log
+  ImmutableAuditLog,
+  AuditEntry,
+  MemoryAuditStore,
+  FileAuditStore,
+  AuditProof,
+  verifyAuditChain,
+
+  // v5.1 — Agent observability exported in Observability section above
+  // v5.1 — Benchmark harness exported in Benchmark Harness section above
+
+  // v6.0 — OWASP LLM Top 10 v2025
+  OWASP_LLM_2025,
+  OWASPCoverageMatrix,
+  OWASP_SEVERITY_WEIGHTS,
+  COVERAGE_MULTIPLIERS,
+
+  // v6.0 — MCP Bridge
+  MCPBridge,
+  MCPToolPolicy,
+  MCPSessionGuard,
+  MCPResourceScanner,
+  MCP_DANGEROUS_TOOLS,
+  ARG_INJECTION_PATTERNS,
+  createMCPMiddleware,
+
+  // v6.0 — NIST AI RMF
+  NIST_AI_RMF_2025,
+  SP800_53_AI_CONTROLS,
+  NISTMapper,
+  AIBOMGenerator,
+  NISTComplianceChecker,
+
+  // v6.0 — EU AI Act
+  EU_AI_ACT_REQUIREMENTS,
+  RiskClassifier,
+  ConformityAssessment,
+  TransparencyReporter,
+  EUIncidentReporter,
+  EUAIActDashboard,
+
+  // v6.0 — System Prompt Leakage (LLM07)
+  PROMPT_EXTRACTION_PATTERNS,
+  SystemPromptGuard,
+  PromptFingerprinter,
+  PromptLeakageMitigation,
+
+  // v6.0 — RAG/Vector Vulnerability (LLM08)
+  RAG_VULNERABILITY_PATTERNS,
+  VECTOR_DB_SECURITY_CHECKLIST,
+  RAGVulnerabilityScanner,
+  EmbeddingIntegrityChecker,
+  RAGPipelineAuditor,
+
+  // Confused Deputy Prevention
+  AuthorizationContext,
+  EphemeralTokenManager,
+  IntentValidator,
+  ConfusedDeputyGuard,
+
+  // v7.0 — MCP SDK Integration
+  shieldMCPServer,
+  createMCPSecurityLayer,
+
+  // v7.0 — MCP Security Runtime
+  MCPSecurityRuntime,
+  MCPSessionStateMachine,
+  SESSION_STATES,
+
+  // v7.0 — MCP Certification & Trust
+  AgentThreatIntelligence,
+  MCPCertification,
+  CrossOrgAgentTrust,
+  MCP_THREAT_CATEGORIES: CERT_THREAT_CATEGORIES,
+  CERTIFICATION_REQUIREMENTS,
+  CERTIFICATION_LEVELS,
+};
+
+// Filter out undefined exports (from modules that failed to load)
+for (const key of Object.keys(_exports)) {
+  if (_exports[key] === undefined) {
+    delete _exports[key];
+  }
+}
+
+module.exports = _exports;