JAEGIS Method Framework - Core Implementation

Framework Overview

The JAEGIS (Brainstorming, Modeling, Architecture, Development) Method is a comprehensive framework for human-AI collaborative software development that emphasizes structured processes, advanced elicitation techniques, and multi-agent coordination.

Core Principles

1. Human-AI Collaboration

• Not "Vibe Coding": Structured approach requiring active human participation

• Collaborative Intelligence: Humans and AI working together to produce better outcomes than either could achieve alone

• Active Engagement: Users don't "check their brain at the door" - continuous collaboration required

2. Multi-Agent Team Approach

• Agile Team of Agents: Multiple specialized agents working together

• Coordinated Workflow: Seamless handoffs between agents and phases

• Holistic Development: End-to-end support from ideation to delivery

3. Advanced Elicitation Techniques

• Beyond Average Responses: 20-30 specialized techniques to push AI beyond generic outputs

• Context-Aware Prompting: Techniques selected based on document section and context

• Psychological-Backed Methods: Science-based approaches to enhance AI reasoning

Core Agents

Primary JAEGIS Agents

brainstorming_agent:
  name: "Brainstorming Specialist"
  role: "Facilitates ideation using psychology-backed techniques"
  capabilities:
    - Science-backed brainstorming methodologies
    - Interactive idea generation
    - Creative problem-solving facilitation
    - Collaborative ideation sessions

product_manager_agent:
  name: "Product Manager (PM)"
  role: "Creates comprehensive Product Requirements Documents"
  capabilities:
    - PRD creation and refinement
    - Stakeholder requirement gathering
    - Feature prioritization
    - User story development

architect_agent:
  name: "System Architect"
  role: "Designs technical architecture and system specifications"
  capabilities:
    - Architecture document creation
    - Technical specification development
    - System design patterns
    - Technology stack recommendations

development_agent:
  name: "Development Coordinator"
  role: "Manages development process and code generation"
  capabilities:
    - Development planning
    - Code generation coordination
    - Quality assurance oversight
    - Deployment preparation

Supporting Agents

elicitation_agent:
  name: "Advanced Elicitation Specialist"
  role: "Applies advanced prompting techniques to enhance AI responses"
  capabilities:
    - 20-30 specialized elicitation techniques
    - Context-aware technique selection
    - Response quality enhancement
    - Multi-perspective analysis

template_agent:
  name: "Agentic Template Manager"
  role: "Manages interactive templates with embedded instructions"
  capabilities:
    - Dynamic template generation
    - Interactive user engagement
    - Context-sensitive questioning
    - Collaborative document creation

expansion_agent:
  name: "Expansion Pack Manager"
  role: "Manages domain-specific expansion packs"
  capabilities:
    - Expansion pack loading and management
    - Domain-specific agent coordination
    - Specialized workflow execution
    - Custom template integration

Workflow Phases

Phase 1: Brainstorming

brainstorming_phase:
  description: "Interactive ideation using psychology-backed techniques"
  agents: [brainstorming_agent, elicitation_agent]
  techniques:
    - Divergent thinking exercises
    - Convergent analysis
    - Multi-perspective simulation
    - "What if" analysis
    - "Yes and" methodology
  outputs:
    - Refined project concept
    - Feature ideas and alternatives
    - Problem-solution mapping
    - Innovation opportunities

Phase 2: Modeling (PRD Creation)

modeling_phase:
  description: "Comprehensive Product Requirements Document creation"
  agents: [product_manager_agent, template_agent, elicitation_agent]
  techniques:
    - Advanced elicitation for requirements gathering
    - Stakeholder simulation
    - Use case development
    - Priority matrix analysis
  outputs:
    - Complete PRD document
    - User stories and acceptance criteria
    - Feature specifications
    - Success metrics definition

Phase 3: Architecture

architecture_phase:
  description: "Technical architecture and system design"
  agents: [architect_agent, template_agent, elicitation_agent]
  techniques:
    - System design elicitation
    - Technology stack analysis
    - Scalability planning
    - Security consideration analysis
  outputs:
    - Architecture document
    - Technical specifications
    - System diagrams
    - Implementation roadmap

Phase 4: Development

development_phase:
  description: "Coordinated development and implementation"
  agents: [development_agent, template_agent]
  techniques:
    - Iterative development planning
    - Code generation coordination
    - Quality assurance integration
    - Deployment preparation
  outputs:
    - Implementation plan
    - Code structure and components
    - Testing strategy
    - Deployment configuration

Advanced Elicitation Techniques Framework

Core Technique Categories

analytical_techniques:
  - "What if analysis"
  - "Root cause analysis"
  - "Pros and cons evaluation"
  - "Risk assessment analysis"
  - "Impact analysis"

creative_techniques:
  - "Yes and methodology"
  - "Alternative generation"
  - "Reverse brainstorming"
  - "Analogical thinking"
  - "Random word association"

collaborative_techniques:
  - "Multiple personality simulation"
  - "Stakeholder perspective taking"
  - "Devil's advocate approach"
  - "Consensus building"
  - "Conflict resolution"

systematic_techniques:
  - "Structured decomposition"
  - "Hierarchical analysis"
  - "Process mapping"
  - "Decision tree analysis"
  - "Priority matrix evaluation"

Technique Selection Logic

technique_selection:
  context_based:
    requirements_gathering: ["stakeholder_simulation", "use_case_analysis", "priority_matrix"]
    architecture_design: ["what_if_analysis", "risk_assessment", "scalability_analysis"]
    problem_solving: ["root_cause_analysis", "alternative_generation", "reverse_brainstorming"]
    creative_ideation: ["yes_and_methodology", "analogical_thinking", "random_association"]
  
  document_section_based:
    introduction: ["context_setting", "stakeholder_identification"]
    requirements: ["elicitation_techniques", "validation_methods"]
    design: ["systematic_analysis", "alternative_evaluation"]
    implementation: ["structured_decomposition", "process_mapping"]

Template System Architecture

Agentic Template Structure

agentic_template:
  metadata:
    template_id: "unique_identifier"
    template_name: "descriptive_name"
    template_type: "prd|architecture|brainstorming|custom"
    version: "semantic_version"
    
  sections:
    - section_id: "unique_section_id"
      section_name: "descriptive_section_name"
      agentic_instructions:
        - instruction_type: "question|brainstorm|analyze|validate"
          instruction_text: "specific_instruction"
          elicitation_techniques: ["technique1", "technique2"]
          expected_interaction: "description_of_expected_user_interaction"
      
  interaction_flow:
    - step: 1
      action: "present_section"
      agent_behavior: "introduce_section_and_ask_initial_questions"
    - step: 2
      action: "elicit_information"
      agent_behavior: "apply_selected_elicitation_techniques"
    - step: 3
      action: "validate_and_refine"
      agent_behavior: "confirm_understanding_and_suggest_improvements"

Template Enhancement Features

template_enhancements:
  dynamic_questioning:
    - Context-aware question generation
    - Follow-up question chains
    - Clarification requests
    - Validation questions
    
  interactive_brainstorming:
    - Real-time idea building
    - Alternative suggestion generation
    - Collaborative refinement
    - Consensus building
    
  adaptive_flow:
    - User response-based navigation
    - Skip logic for completed sections
    - Conditional section activation
    - Progress tracking and resumption

Expansion Pack System

Expansion Pack Architecture

expansion_pack:
  metadata:
    pack_id: "unique_pack_identifier"
    pack_name: "descriptive_pack_name"
    domain: "target_domain" # e.g., "unity_2d_games", "web_development"
    version: "semantic_version"
    author: "pack_creator"
    
  specialized_agents:
    - agent_id: "domain_specific_agent_id"
      agent_name: "descriptive_agent_name"
      specialization: "specific_domain_expertise"
      capabilities: ["capability1", "capability2"]
      
  custom_templates:
    - template_id: "domain_specific_template_id"
      template_name: "descriptive_template_name"
      template_type: "domain_specific_type"
      
  workflow_modifications:
    - phase: "brainstorming|modeling|architecture|development"
      modifications: ["modification_description"]
      additional_steps: ["step_description"]

Example Expansion Packs

unity_2d_game_pack:
  domain: "Unity 2D Game Development"
  specialized_agents:
    - "Unity Game Designer"
    - "2D Asset Coordinator"
    - "Game Mechanics Specialist"
  custom_templates:
    - "Game Design Document Template"
    - "Unity Project Architecture Template"
    - "2D Asset Pipeline Template"

web_development_pack:
  domain: "Modern Web Development"
  specialized_agents:
    - "Frontend Architecture Specialist"
    - "Backend API Designer"
    - "Full-Stack Integration Coordinator"
  custom_templates:
    - "Web Application PRD Template"
    - "Modern Web Architecture Template"
    - "API Specification Template"

Integration Points

Platform Integration Framework

platform_integrations:
  claude_code:
    integration_type: "native"
    features: ["full_workflow", "template_support", "agent_coordination"]
    setup_requirements: ["claude_code_access", "jaegis_configuration"]
    
  gemini_gems:
    integration_type: "gems_based"
    features: ["agent_gems", "workflow_gems", "template_gems"]
    setup_requirements: ["gemini_access", "gems_development_kit"]
    
  chatgpt_custom_gpts:
    integration_type: "custom_gpt"
    features: ["agent_gpts", "workflow_coordination", "cross_gpt_communication"]
    setup_requirements: ["chatgpt_plus", "custom_gpt_creation_access"]

Cross-Platform Synchronization

synchronization:
  project_data:
    - Project metadata and configuration
    - Document versions and history
    - Agent interaction logs
    - Template customizations
    
  user_preferences:
    - Preferred elicitation techniques
    - Template customizations
    - Workflow preferences
    - Agent interaction styles

This foundational framework provides the structure for implementing all the advanced JAEGIS method features identified from Brian's presentation, with clear separation of concerns and extensibility for future enhancements.