mindXdashboard/docs/book/journal/api/dojo/inference/governance/origin
philosophymanifestothesisoriginwhitepaperataraxiaroadmappress|archovervieworchestrationcodebasehierarchycore|agentsmindXagentceomastermindbdievolutionauthorall
govdaiocivilizationidentitysecurity|memorypgvectorembedaglmmemory|inferencevllmollamamistralgemini|timeoracle
toolsindextoolsa2amcpshell|publishauthoragentbookjournal|deployproductionsecuritymonitoring|apireferenceswagger|learnusageguidehackathon
AGENTS.md · 33.9 KB

MindX Agent Architecture Reference

Comprehensive Agent Registry & Documentation

Version: 4.0 Date: 2026-03-31 Status: ✅ Production Ready - Enterprise deployment with encrypted vault security Purpose: Complete reference for all agents in the mindX orchestration environment

🎯 Agent Architecture Overview

mindX operates as an agnostic orchestration environment with a multi-tier agent architecture supporting symphonic coordination across intelligence levels. Each agent maintains cryptographic identity, specialized capabilities, and standardized A2A (Agent-to-Agent) communication.

🎼 Orchestration Hierarchy

Higher Intelligence → CEO.Agent → Conductor.Agent → mindX Environment (CORE)
                                                        ↓
                                              MastermindAgent (Coordinator)
                                                   ↓     ↓
                                    Specialized Agents  AION Agent (Autonomous)
                                                        ↓
                                                 SystemAdmin Agent
                                                 Backup Agent
                                                 Chroot Environments

🔗 AION Containment & Autonomy Model

AION operates within a sophisticated containment hierarchy while maintaining operational autonomy:

Containment Structure:

Autonomy Levels:

  1. COMPLY: Execute directive as specified
  2. REFUSE: Reject directive with reasoning
  3. MODIFY: Execute modified version of directive
  4. DEFER: Delay execution pending conditions
  5. AUTONOMOUS: Independent action without directive

Control Flow: 

CORE System → MastermindAgent → Directive → AION Agent → Decision Logic
                                              ↓
                                        [COMPLY|REFUSE|MODIFY|DEFER|AUTONOMOUS]
                                              ↓
                                     SystemAdmin Agent + AION.sh Execution

🎯 Executive Summary

The mindX orchestration environment is now PRODUCTION READY with enterprise-grade security infrastructure, encrypted vault management, and complete cryptographic identity system for all agents and tools.

🚀 Production-Ready Infrastructure

🧠 CORE System Analysis

🔐 Security Enhancements

📋 COMPLETE AGENT REGISTRY (CORE + Specialized)

🧠 CORE AGENTS (15 Foundational Components)

These are the foundational agents that provide cognitive, infrastructure, and orchestration capabilities for the entire mindX system. ALL other agents depend on CORE.

🎯 Tier 1: Meta-Orchestration

🌟 MindXAgent (mindx_meta_orchestrator)

- agent_knowledge: Comprehensive knowledge base of all agents - agent_capabilities: Dynamic capability mapping and analysis - improvement_goals: Self-improvement target management - orchestrate_improvement(): Coordinate system-wide improvements - analyze_agent_capabilities(): Deep understanding of agent relationships

🧠 Tier 2: Cognitive Architecture

🎯 BDIAgent (bdi_reasoning_core)

- belief_system: Shared singleton belief management - execute_tool(): Context-aware tool execution - failure_analyzer: Intelligent error classification and recovery - generate_plan(): Multi-step action planning

🎭 AGInt (agint_cognitive_coordinator)

- run_poda_loop(): Perception → Orientation → Decision → Action - process_primary_directive(): Execute high-level objectives - stuck_loop_detector: Infinite loop prevention - exit_detector: Completion condition detection

🧩 BeliefSystem (singleton_belief_manager)

- beliefs: Dict[str, Belief] with confidence scoring - update_belief(): Thread-safe belief updates - query_beliefs(): Context-aware belief retrieval - Source tracking (PERCEPTION, INFERENCE, LEARNED, etc.)

🔧 Tier 3: Infrastructure Services

🎼 CoordinatorAgent (central_service_bus)

- interactions: Request/response tracking - subscribers: Event pub/sub system - health_status: System health metrics aggregation - route_interaction(): Intelligent request routing - publish_event(): Event broadcasting

💾 MemoryAgent (persistent_memory_infrastructure)

- save_timestamped_memory(): Store timestamped records - promote_stm_to_ltm(): Memory promotion based on importance - analyze_agent_patterns(): Behavioral pattern extraction - get_agent_memory_context(): Context retrieval for tasks

🆔 IDManagerAgent (cryptographic_identity_ledger)

- create_new_wallet(): Generate Ethereum-compatible wallets - store_identity(): Maintain cryptographic identity records - map_entity_to_address(): Entity ↔ address mapping - sign_message(): Cryptographic message signing

🛡️ GuardianAgent (security_infrastructure)

- Challenge-response authentication system - Private key access arbitration - Security validation and audit logging - Agent registration verification

🎭 Tier 4: Strategic Orchestration

🎭 MastermindAgent (strategic_coordinator)

- Strategic planning and goal setting - Campaign orchestration and management - High-level objective coordination - AION agent directive generation and management

🚀 StartupAgent (system_bootstrap_controller)

- bootstrap_system(): Complete system initialization - Dependency resolution and startup ordering - Agent registry loading and validation - Configuration and environment setup

🛠️ Tier 5: Core Utilities

🧮 ReasoningAgent (reasoning_engine)

🎓 EpistemicAgent (knowledge_manager)

🔄 NonMonotonicAgent (belief_revision)

📋 SessionManager (session_lifecycle)

🔍 StuckLoopDetector & ExitDetector

🚀 SPECIALIZED AGENTS (Built on CORE Foundation)

These agents provide domain-specific capabilities and depend on the CORE infrastructure for their operation.

🎯 Tier 1: Strategic Services

🛡️ GuardianAgent (guardian_agent_main)

🆔 IDManagerAgent (default_identity_manager)

Tier 3: Specialized Services

🧭 AGInt (agint_coordinator)

🎯 BDIAgent (bdi_agent_mastermind_strategy)

🤖 AutoMINDXAgent (automindx_agent_main)

🧬 StrategicEvolutionAgent (sea_for_mastermind)

🏗️ BlueprintAgent (blueprint_agent_mindx_v2)

🧬 StrategicEvolutionAgent (strategic_evolution_coordinator)

🎯 Tier 2: Autonomous Operations

⚡ AION Agent (aion_prime)

- CORE-contained: Operates within CORE orchestration layer - MASTERMIND-directed: Receives strategic directives from MastermindAgent (CORE) - Autonomous: Maintains decision sovereignty (comply/refuse/modify/defer) - Chroot environment creation, optimization, and migration - mindX system replication across environments - Autonomous decision-making with sovereignty levels (1-5) - Cross-environment operations and maintenance

🔧 SystemAdminAgent (systemadmin_for_aion)

💾 BackupAgent (backup_agent_main)

🔧 Tier 3: Development & Analysis

🖥️ EnhancedSimpleCoder + SimpleCoder

📊 AnalyzerAgent + BenchmarkAgent

🎭 Tier 4: Persona & Identity

🎨 PersonaAgent + AvatarAgent

🤖 AutoMINDXAgent (automindx_agent_main)

📊 Tier 5: Monitoring & Health

📈 PerformanceMonitor + ResourceMonitor

📊 CORE vs SPECIALIZED CLASSIFICATION

✅ CORE Components (15 Foundational)

Required for basic system function - cannot be removed:

ComponentSizeCriticalityCORE Function
MindXAgent~149KBCRITICALMeta-orchestration
BDIAgent~64KBCRITICALReasoning engine
CoordinatorAgent~56KBCRITICALService bus
MemoryAgent~53KBCRITICALMemory infrastructure
StartupAgent~83KBCRITICALSystem bootstrap
MastermindAgent~41KBHIGHStrategic control
AGInt~32KBHIGHCognitive loop
IDManagerAgent~16KBHIGHIdentity management
GuardianAgent~16KBHIGHSecurity infrastructure
BeliefSystem~8KBCRITICALShared knowledge
+ 5 Core Utilities~60KBMEDIUM-HIGHSupport functions

Total CORE: ~582KB of foundational cognitive and infrastructure code

❌ Specialized Components (25+ Domain-Specific)

Built on CORE foundation - can be added/removed based on needs:

⚠️ UNREGISTERED AGENTS (Found in Codebase)

Critical Core Agents

🧠 EnhancedMemoryAgent

🧠 MemoryAgent

📈 SelfImprovementAgent

Utility & Development Agents

💻 EnhancedSimpleCoder

🌐 MultiModelAgent

📚 DocumentationAgent

Tier 4: Cognitive Sub-Agents (Augmentic Development)

🔮 PredictionAgent (prediction_agent)

🧠 ReasoningAgent (reasoning_agent)

🔄 NonMonotonicAgent (nonmonotonic_agent)

📚 EpistemicAgent (epistemic_agent)

❓ SocraticAgent (socratic_agent)

Tier 5: Lifecycle Management Agents

🚀 StartupAgent (startup_agent)

🔄 ReplicationAgent (replication_agent)

🛑 ShutdownAgent (shutdown_agent)

Tier 6: Storage & Registry Agents

⛓️ BlockchainAgent (blockchain_agent)

Testing & Validation Agents

🧪 UltimateCognitionTestAgent

🧪 EnhancedUltimateCognitionTestAgent

📊 ReportAgent

🚨 CRITICAL REGISTRATION GAPS

Immediate Action Required

  1. IDManagerAgent - Core identity management (CRITICAL)
  2. AGInt - Central cognitive engine (CRITICAL)
  3. BDIAgent - Primary execution engine (CRITICAL)
  4. EnhancedMemoryAgent - Advanced memory capabilities (HIGH)
  5. SelfImprovementAgent - Autonomous improvement (HIGH)

Agent Registration Priority Matrix

AgentPriorityReasonImpact
IDManagerAgentCRITICALIdentity management for all agentsSystem failure
AGIntCRITICALCore cognitive processingNo intelligence
BDIAgentCRITICALPrimary execution engineNo task execution
EnhancedMemoryAgentHIGHAdvanced learning capabilitiesLimited learning
SelfImprovementAgentHIGHAutonomous evolutionNo self-improvement
MultiModelAgentMEDIUMLLM optimizationPerformance impact
SimpleCoderAgentMEDIUMDevelopment capabilitiesLimited coding

📊 Agent Statistics

- PredictionAgent, ReasoningAgent, NonMonotonicAgent, EpistemicAgent, SocraticAgent - StartupAgent, ReplicationAgent, ShutdownAgent, BlockchainAgent

🔧 Agent Architecture Principles

Core Design Patterns

  1. Cryptographic Identity: Every agent has unique identity and signature
  2. A2A Communication: Standardized agent-to-agent protocols
  3. Tool Integration: Agents use tools, not internal implementations
  4. Memory-Driven Learning: Comprehensive logging enables evolution
  5. BDI Cognitive Model: Belief-Desire-Intention for goal-oriented behavior

Registration Requirements

🎯 Next Steps

  1. Register Critical Agents: IDManager, AGInt, BDI immediately
  2. Audit Agent Capabilities: Ensure all agents have proper tool access
  3. Standardize A2A Communication: Implement model cards for all agents
  4. Memory Integration: Connect all agents to enhanced memory system
  5. Performance Monitoring: Track agent effectiveness and resource usage

🔐 ENHANCED SECURITY ARCHITECTURE

Identity Management Workflow

1. Agent Creation → IDManager.create_new_wallet()

  1. Registry Validation → Guardian.validate_registry_status()
    2. 

  2. Identity Verification → Guardian.validate_identity()
    3. 

  3. Challenge-Response → Guardian.challenge_response_test()
    4. 

  4. Workspace Validation → Guardian.validate_workspace()
    5. 

  5. Production Approval → Guardian.approve_agent_for_production()
    6.

Tool Security Features

🎯 NEXT PHASE PRIORITIES

Phase 2: Complete Agent Registration

  1. MemoryAgent & EnhancedMemoryAgent - Memory management
  2. SimpleCoderAgent - Development capabilities
  3. MultiModelAgent - LLM optimization
  4. TestAgent & ReportAgent - Testing infrastructure

Phase 3: Advanced Security

🏆 IDENTITY MANAGEMENT ACHIEVEMENT

The mindX system has successfully implemented enterprise-grade identity management with:

This represents a MAJOR MILESTONE in the mindX orchestration environment's evolution toward a fully autonomous, secure, and trusted multi-agent system.

This documentation reflects the current state after the IDENTITY MANAGEMENT OVERHAUL and highlights the significant security improvements achieved.*

All DocumentsDocument IndexThe Book of mindXImprovement JournalAPI Reference