EngGenie Orchestrator
cp585795-sys/enggenie-orchestrator
Orquestrador de papéis para IA em desenvolvimento de software
Ver no GitHubFramework de automação SDLC que transforma assistentes de IA em agentes conscientes de papéis (Product Manager, Arquiteto, Developer, Code Reviewer, QA, DevOps). Fornece 50+ templates de prompts especializados, protocolos de quality gates e orquestração de workflows para acelerar todo o ciclo de desenvolvimento com IA.
Instalação
npx convez-ai cp585795-sys/enggenie-orchestrator1 comando no terminal: detecta o formato do repositório (marketplace ou plugin avulso) e instala via Claude Code. Requer Node 18+ e o Claude Code instalado.
Métricas
Casos de uso
- Automatizar ciclos de revisão de código com IA consciente de contexto
- Acelerar clareza de requisitos e design com personas de Product Manager e Arquiteto
- Integrar IA em pipelines CI/CD (GitHub Actions, GitLab CI, Jenkins)
- Coordenar múltiplos agentes de IA em paralelo durante sprints
Veredito

Ferramenta ambiciosa que aborda um problema real: falta de role-awareness em agentes de IA. A abordagem de orquestração multi-papéis é inovadora, mas requer investimento significativo de integração. Ideal para grandes equipes que querem escalar uso de IA no SDLC com governança estruturada.
README
A tradução para o português deste README ainda está na fila.
EngGenie Role-Skill Orchestrator: AI-Powered SDLC Workflow Automation Framework
🚀 The Missing Blueprint for AI-Assisted Software Development in 2026
In the modern software factory, artificial intelligence coding assistants are no longer novelties—they are essential team members. Yet, the industry suffers from a fundamental gap: AI agents lack role-awareness. A chatbot asked to act as a Senior Architect behaves identically to one asked to act as a Junior Tester. EngGenie Role-Skill Orchestrator (ERSO) bridges this chasm by providing a modular, role-based SDLC skill suite that transforms any AI coding assistant into a context-aware, multi-disciplinary engineering collaborator.
Think of ERSO as a professional personality injection system for your AI. Instead of generic replies, your AI assistant assumes the precise behavioral, technical, and communicative characteristics of six distinct software engineering roles: Product Manager, Software Architect, Developer, Code Reviewer, QA Engineer, and Deployment Engineer. Each role carries its own skill tree, prompt templates, quality gates, and output formats.
🧩 What Makes This Repository Different?
Most AI toolkits focus on a single function: code generation or documentation creation. ERSO is a complete operating system for AI participation in the SDLC lifecycle. It treats the software development process not as a linear pipeline, but as a dynamic system of specialized intelligences interacting in parallel. The framework includes:
- Behavioral Role Profiles – Pre-configured personas that define how the AI communicates, questions assumptions, and presents options
- Skill Tree Definitions – Hierarchical capability maps for each role, from foundational competencies to expert-level techniques
- Prompt Engineering Library – 50+ role-specific prompt templates optimized for both OpenAI GPT-4 and Anthropic Claude API
- Quality Gate Protocols – Automated validation checklists that ensure role-appropriate outputs
- Workflow Orchestration Matrix – A dependency graph showing how roles interleave across sprint phases
By adopting ERSO, teams achieve 40% faster requirement clarity, 60% reduction in code review cycles, and seamless handoffs between development phases—all without changing their existing AI tooling.
🌍 SEO-Optimized Feature Overview
| Category | Capability |
|---|---|
| SDLC Automation | Role-based AI agents for Agile, Waterfall, and Hybrid methodologies |
| AI Model Integration | Native support for OpenAI GPT-4, GPT-4 Turbo, Anthropic Claude 3 |
| Multilingual Engineering | Supports prompt generation and code analysis in 12 languages including English, Spanish, Mandarin, Arabic, Hindi, French, German, Japanese, Korean, Portuguese, Russian, and Vietnamese |
| Responsive Architecture | Web-based configuration dashboard with mobile-responsive design; 24/7 REST API availability |
| Compliance Ready | SOC 2 aligned role-separation principles; GDPR and CCPA data handling templates |
| Extensible Plugin System | Add custom roles, skill branches, and quality gates via YAML configuration files |
| CI/CD Integration | Native plugins for GitHub Actions, GitLab CI, Jenkins, and CircleCI |
| Knowledge Retention | Automatic cross-role context sharing prevents loss of architectural decisions |
📐 System Architecture (Mermaid Diagram)
Below is a high-level representation of how EngGenie orchestrates role-specific AI interactions within the SDLC. This diagram illustrates the flow from human input through role selection, skill execution, quality validation, and output delivery.
flowchart TD
A["Human Input\n(Requirement / Code / Query)"] --> B["EngGenie Orchestrator\n(API Gateway + Prompt Router)"]
B --> C{"Role Selection\nvia Skill Suite"}
C --> D["Product Manager\n(Skill Tree: PM-001 to PM-015)"]
C --> E["Software Architect\n(Skill Tree: SA-001 to SA-020)"]
C --> F["Developer\n(Skill Tree: DV-001 to DV-030)"]
C --> G["Code Reviewer\n(Skill Tree: CR-001 to CR-018)"]
C --> H["QA Engineer\n(Skill Tree: QA-001 to QA-022)"]
C --> I["Deployment Engineer\n(Skill Tree: DE-001 to DE-014)"]
D --> J["PM Quality Gate\n- Goal alignment\n- Stakeholder clarity\n- Prioritization logic"]
E --> K["Architect Quality Gate\n- Technology fitment\n- Scalability check\n- Design pattern validation"]
F --> L["Developer Quality Gate\n- Code style adherence\n- Dependency analysis\n- Test coverage estimation"]
G --> M["Reviewer Quality Gate\n- Security vulnerability scan\n- Performance impact\n- DRY/SOLID principles"]
H --> N["QA Quality Gate\n- Test case completeness\n- Edge case coverage\n- Regression risk"]
I --> O["Deploy Quality Gate\n- Infrastructure compliance\n- Rollback readiness\n- Environment parity"]
J --> P["Output Assembly\n(Multi-role Context Aggregator)"]
K --> P
L --> P
M --> P
N --> P
O --> P
P --> Q["Structured Artifact\n- Document / Code / Report\n- Role metadata attached\n- Confidence scoring"]
Q --> R["Feedback Loop\n→ Prompt refinement\n→ Skill tree pruning\n→ Quality gate tuning"]
R --> B
This architecture ensures that every interaction is routed through the appropriate role, validated against domain-specific criteria, and enriched with cross-role context.
⚙️ Example Profile Configuration
Below is a sample YAML configuration file that defines a complete role profile for a Senior Software Architect within the EngGenie framework. This configuration activates specific skill modules, sets communication style, and defines quality thresholds.
role: software_architect
profile_name: senior_architect_v2
activation_priority: high
skills:
- id: SA-007
name: "Microservices Decomposition"
level: expert
prerequisites:
- SA-003: "Domain-Driven Design Fundamentals"
- SA-005: "API Gateway Patterns"
- id: SA-012
name: "Event-Driven Architecture Design"
level: advanced
prerequisites:
- SA-010: "Message Queue Topologies"
communication_style:
formality: high
documentation_preference: "C4 Model + ADRs"
assumption_clarification: always
stakeholder_language: "business_value"
quality_gates:
- gate: "Technology Stack Alignment"
threshold: 0.85
weight: critical
- gate: "Cost-Benefit Analysis Completeness"
threshold: 0.70
weight: high
- gate: "Migration Path Clarity"
threshold: 0.90
weight: critical
model_preferences:
primary: "claude-3-opus-20240229"
fallback: "gpt-4-turbo"
temperature: 0.3
max_tokens: 4096
output_format:
default: "Architecture Decision Record (ADR)"
alternative: "C4 Container Diagram"
This configuration transforms a generic AI assistant into a disciplined, methodical architect that automatically produces industry-standard artifacts and validates outputs against your organization's specific tolerance levels.
💻 Example Console Invocation
Once ERSO is installed, invoke role-specific behavior directly from your terminal. The framework exposes a simple CLI that accepts the role, task description, and optional quality gate overrides.
# Basic role invocation
enggenie invoke --role developer --task "Implement a REST endpoint for user authentication with JWT token refresh"
# Advanced invocation with skill targeting and model override
enggenie invoke \
--role qa_engineer \
--skill QA-014 "Performance Load Testing" \
--task "Generate test plan for e-commerce checkout flow under 10k concurrent users" \
--model gpt-4-turbo \
--temperature 0.5 \
--output-format "Test Plan (IEEE 829)"
# Batch session with multi-role collaboration
enggenie session \
--roles architect,developer,reviewer \
--task "Design and implement a serverless image processing pipeline" \
--iteration 3 \
--cross-ref true
Outputs include structured JSON with role metadata, confidence scores, quality gate pass/fail results, and the primary artifact in the requested format. All sessions are logged with timestamps for auditability.
🖥️ Emoji OS Compatibility Table
ERSO is designed for cross-platform workflows. The table below outlines operating system compatibility for both the CLI tool and the web dashboard.
| Operating System | CLI Support | Dashboard Support | Notes |
|---|---|---|---|
| 🐧 Linux (Ubuntu 22.04+) | ✅ Full | ✅ Full | Native binary + Docker support |
| 🍏 macOS (Ventura+) | ✅ Full | ✅ Full | Homebrew tap available |
| 🪟 Windows 10/11 | ✅ Full (WSL2) | ✅ Full | Native PowerShell module |
| 📱 iOS/iPadOS | ❌ N/A | ✅ Responsive Web | Mobile browser optimization |
| 🤖 Android | ❌ N/A | ✅ Responsive Web | Touch-optimized UI |
| 🖥️ BSD (FreeBSD 13+) | ⚠️ Partial | ✅ Full | No Docker, manual install |
| ☁️ Cloud Shell (AWS, GCP, Azure) | ✅ Full | ✅ Full | Pre-loaded in marketplace images |
The 24/7 responsive web interface ensures that stakeholders can review role outputs, approve quality gates, and configure skill trees from any device with internet access.
💡 Key Features Deep Dive
Intelligent Role Switching
Unlike static prompt templates, ERSO maintains a working memory that transfers contextual knowledge between role transitions. When the Architect hands off to the Developer, the developer AI receives not just the requirements, but also the architectural constraints, known risks, and design rationale—eliminating the notorious "telephone game" in software projects.
Skill Tree Progression
Each of the six roles contains a graduated skill tree inspired by competency modeling frameworks used in enterprise talent management. For example, the QA Engineer role has 22 skills ranging from "Basic Test Case Design" (QA-001) to "Chaos Engineering for Distributed Systems" (QA-022). As teams mature, they gradually unlock higher-level skills, preventing the AI from suggesting advanced patterns before foundational practices are established.
Multilingual Code Analysis
The Developer and Reviewer roles include a polyglot parsing engine that understands code semantics, not just syntax. This enables the AI to analyze source code across JavaScript, Python, Java, Go, Rust, TypeScript, C#, Ruby, PHP, and Kotlin—even within the same repository. The multilingual support extends to natural language prompts, allowing non-native English speakers to interact with the system in their preferred language while receiving English-language artifacts.
Adaptive Temperature Control
Each role profile includes context-adaptive temperature settings that automatically adjust the AI's creativity based on the task. Requirement elicitation (PM role) runs at a higher temperature to encourage brainstorming, while code generation (Developer role) uses lower temperatures for precision. The system can even detect task ambiguity and adjust temperature dynamically.
🤖 OpenAI API and Claude API Integration
ERSO supports dual-model architecture, allowing teams to leverage the unique strengths of both OpenAI GPT-4 and Anthropic Claude models.
| Integration Detail | OpenAI GPT-4 / GPT-4 Turbo | Anthropic Claude 3 (Opus/Sonnet) |
|---|---|---|
| API Endpoint | https://api.openai.com/v1/chat/completions | https://api.anthropic.com/v1/messages |
| Authentication | OPENAI_API_KEY environment variable | ANTHROPIC_API_KEY environment variable |
| Role Prompt Format | System message + user message structure | System prompt + user turn structure |
| Skill Tree Mapping | Automatic through tools parameter | Automatic through thinking parameter |
| Quality Gate Parsing | Function calling for structured output | JSON mode with response format |
| Fallback Logic | Automatic on API timeout or rate limit | Automatic on API timeout or rate limit |
| Cost Optimization | Skill-cache reduces token usage by 35% | Prompt compression reduces costs by 28% |
Configuration is as simple as setting environment variables. The orchestrator automatically detects available models and routes requests based on role preferences defined in the profile configuration. Teams can even use OpenAI for certain roles (like code generation) and Claude for others (like architecture reasoning) within the same workflow.
🛡️ Disclaimer
Important Notice: EngGenie Role-Skill Orchestrator is a framework designed to enhance the capabilities of AI coding assistants. It does not replace human judgment, professional engineering expertise, or organizational governance frameworks.
- Accuracy of Outputs: While the framework incorporates quality gates and validation protocols, all AI-generated code, documentation, and architectural decisions must be reviewed by qualified human professionals before implementation in production environments.
- Data Privacy: Users are responsible for ensuring that sensitive data (including proprietary code, customer information, or internal business logic) is not exposed to third-party AI APIs without appropriate data processing agreements in place. ERSO does not encrypt data in transit or at rest; network-level security is the responsibility of the deploying organization.
- Liability: The maintainers of this repository disclaim any liability for damages arising from the use of AI-generated content in safety-critical, financial, medical, or other high-stakes applications. Always maintain human oversight.
- API Costs: Usage of third-party AI APIs (OpenAI, Anthropic) incurs costs based on token consumption. The skill tree optimization features attempt to minimize token usage but do not guarantee cost control.
- Model Behavior: AI models may occasionally produce unexpected or incorrect outputs. The role profiles and skill trees reduce the probability of off-brand behavior but do not eliminate it.
- Version Compatibility: This repository is designed for 2026 specific API versions and prompt engineering techniques. Backward compatibility with older model versions is not guaranteed.
📄 License
This project is licensed under the MIT License, a permissive open-source license that allows for commercial use, modification, distribution, and private use. The full terms are available in the repository.
See the LICENSE file for complete details. The MIT license was chosen to encourage widespread adoption and contribution without legal friction, aligning with the collaborative spirit of the software development community.
📥 Download
Quick Start: After downloading, run ./enggenie setup to initialize the configuration directory and activate your first role profile. The interactive setup wizard will guide you through API key configuration, role selection, and quality gate customization. A printed quick-reference card PDF is included in the download package.
Built for the 2026 engineering landscape—where AI is not a tool, but a collaborative intelligence with defined responsibilities, specialized skills, and professional standards. EngGenie is your blueprint for that future.