AEGIS OS

AEGIS OS (Artificial Executive Governance & Intelligence System) is a multi-agent AI orchestration framework that integrates 6 executive organizations and 24 specialized AI agents. Operating as a Claude Code submodule, it launches your entire AI team with a single command.

Key Features

Multi-Agent Orchestration

Integrates 6 executive organizations and 24 AI agents into a single framework. Each agent has a specialized domain, and the right agent activates with a single command.

Executive Agents

CommandAgentRole
/ceoAI_CEOStrategic decisions
/ctoAI_CTOTechnical architecture
/cfoAI_CFOFinancial analysis
/croAI_CRORisk assessment
/cmoAI_CMOMarketing strategy
/cpoAI_CPOProduct management
/cdoAI_CDODesign leadership

Specialist Layer

CommandAgentDomain
/implImplementationAgentCode implementation & feature development
/qaQA_LeadQuality assurance & test strategy
/secSecurityReviewerSecurity review
/techTechnicalReviewerTechnical review
/designDesignerDesign tasks
/uxUX_DesignerUX design
/uiUI_DesignerUI design

Red Team Layer

CommandAgentDomain
/advocateDevil’s AdvocateOpposing views & assumption testing
/skepticSkepticAlternative analysis & skeptical perspective

Support Layer

CommandAgentDomain
/investorAI_InvestorInvestment analysis
/researchResearcherResearch
/dataDataScientistData analysis
/infraInfraEngineerInfrastructure design
/promptPromptEngineerPrompt design
/growthGrowth_HackerGrowth strategy
/hrHR_ManagerAgent management
/ethicsEthicsReviewerEthics review

Persona System

11 specialized personas auto-activate based on context.

  • Architect — Systems design & long-term architecture
  • Frontend — UI/UX & accessibility
  • Backend — Server-side & reliability engineering
  • Security — Threat modeling & vulnerability assessment
  • Performance — Optimization & bottleneck elimination
  • Analyzer — Root cause analysis & investigation
  • QA — Quality assurance & testing
  • Refactorer — Code quality & tech debt management
  • DevOps — Infrastructure & deployment automation
  • Mentor — Education & knowledge transfer
  • Scribe — Documentation & localization

Wave Execution Engine

An engine that automatically decomposes complex tasks into multiple waves for staged execution.

Execution Strategies:

  • Progressive — Incremental enhancement
  • Systematic — Comprehensive methodical analysis
  • Adaptive — Dynamic configuration
  • Enterprise — Large-scale orchestration

MCP Integration

Integrates 4 MCP servers to extend Claude Code capabilities.

ServerPurpose
Context7Library docs & framework patterns
SequentialComplex analysis & multi-step reasoning
MagicUI component generation & design systems
PlaywrightE2E testing & performance metrics

Multi-LLM Orchestration

Unified utilization of Claude, Gemini, and Codex.

  • /multi-llm — Claude as CEO coordinates Gemini and Codex
  • /review-llm — Cross-review by 3 LLMs
  • /codex — Delegate tasks to OpenAI Codex
  • /gemini — Delegate tasks to Google Gemini

Red Team Analysis

Built-in adversarial analysis to improve decision quality.

  • /advocate — Devil’s Advocate presents opposing views
  • /skeptic — Skeptic provides alternative analysis
  • /redteam — Both Red Team agents for comprehensive critical review
  • /boardroom — CEO-led all-agent boardroom debate

Getting Started

Prerequisites

Installation

# 1. Clone the repository
git clone https://github.com/pyonkichi369/aegis-os.git

# 2. Integrate into your project (as submodule)
cd your-project
git submodule add https://github.com/pyonkichi369/aegis-os.git aegis

# 3. Start using AEGIS commands
/aegis-os          # Run full organization pipeline
/ceo               # Strategic decision making
/boardroom         # All-agent boardroom debate