cp-toolkit 2.2.10 → 2.2.12

package/README.md

@@ -124,6 +124,79 @@ applyTo: "**/*.ts,**/*.tsx"
 | `.github/instructions/*.instructions.md` | Path-specific rules |
 | `AGENTS.md` | Universal AI instructions |
 | `.vscode/mcp.json` | MCP server configuration |
+| `.github/cp-kit-models.yaml` | AI model allocation matrix |
+
+## Architect-Builder Strategy
+
+cp-toolkit implements the **Architect-Builder Pattern** for optimal AI agent performance. This strategy separates reasoning from execution:
+
+### Dual-Mode Architecture
+
+```
+┌─────────────────────────────────────────────────────────────────┐
+│  SINGLE MODE                    HYBRID MODE (Architect-Builder) │
+│  ─────────────                  ───────────────────────────────  │
+│                                                                  │
+│  ┌─────────────────┐            PLANNER         EXECUTOR         │
+│  │ High-IQ Model   │            (Architect)     (Builder)        │
+│  │ Pure Reasoning  │            ┌──────────┐    ┌──────────┐     │
+│  └─────────────────┘            │ temp 0.1 │ →  │ temp 0.3 │     │
+│                                 │ Strategy │    │ Code Gen │     │
+│  • orchestrator                 └──────────┘    └──────────┘     │
+│  • security-auditor                                              │
+│  • debugger                     • backend-specialist             │
+│  • documentation-writer         • frontend-specialist            │
+│                                 • devops-engineer                │
+└─────────────────────────────────────────────────────────────────┘
+```
+
+### Mode Selection
+
+| Mode | Temperature | Use Case |
+|------|-------------|----------|
+| **Single** | 0.1 | Pure reasoning tasks (analysis, planning, auditing) |
+| **Hybrid Planner** | 0.1 | Strategic thinking, architecture decisions |
+| **Hybrid Executor** | 0.3 | Code generation, implementation |
+
+### Agent Categories
+
+1. **Leadership & Strategy** (Single Mode)
+   - `orchestrator`, `product-manager`, `product-owner`, `project-planner`
+   
+2. **Development Core** (Hybrid Mode)
+   - `backend-specialist`, `frontend-specialist`, `mobile-developer`, `game-developer`
+
+3. **Infrastructure & Ops** (Hybrid Mode)
+   - `devops-engineer`, `database-architect`, `security-auditor`, `penetration-tester`
+
+4. **Quality & Optimization** (Hybrid Mode)
+   - `qa-automation-engineer`, `test-engineer`, `performance-optimizer`, `debugger`
+
+5. **Specialists & Research** (Mixed)
+   - `code-archaeologist`, `documentation-writer`, `seo-specialist`, `explorer-agent`
+
+### Why This Works
+
+- **Planner (Architect)**: Uses high-reasoning models with low temperature for accurate strategic decisions
+- **Executor (Builder)**: Uses fast code-generation models with slightly higher temperature for creative implementation
+- **Cost Optimization**: Expensive models only for planning; economical models for bulk code generation
+- **Quality Assurance**: Clear separation prevents "hallucination drift" in long implementations
+
+### Models Configuration
+
+The `cp-kit-models.yaml` file defines the model allocation:
+
+```yaml
+agents:
+  backend-specialist:
+    mode: "hybrid"
+    planner:
+      model: "gpt-5.2"
+      task: "API architecture, security and data modeling"
+    executor:
+      model: "gpt-5.2-codex"
+      task: "Route and service implementation with perfect typing"
+```
 
 ## License
 

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "cp-toolkit",
-  "version": "2.2.10",
+  "version": "2.2.12",
   "description": "Copilot Toolkit - AI Agent framework for GitHub Copilot, Claude, Gemini CLI, and other AI assistants",
   "keywords": [
     "ai-agents",

package/src/commands/init.js

@@ -119,8 +119,8 @@ export async function initCommand(directory, options) {
     // 5. Setup .github/copilot-instructions.md
     spinner.text = 'Creating copilot-instructions.md...';
     const instructionsPath = path.join(targetDir, '.github', 'copilot-instructions.md');
-    const instructionsContent = generateCopilotInstructions(config);
-    await fs.writeFile(instructionsPath, instructionsContent);
+    const instructionsTemplatePath = path.join(templatesDir, 'copilot-instructions.md');
+    await fs.copy(instructionsTemplatePath, instructionsPath);
 
     // 6. Setup .github/cp-kit-models.yaml
     spinner.text = 'Creating cp-kit-models.yaml...';
@@ -689,74 +689,6 @@ applyTo: ".github/workflows/**/*.yml,.github/workflows/**/*.yaml"
   }
 }
 
-function generateCopilotInstructions(config) {
-  return `# GitHub Copilot Instructions
-
-> **Copilot Kit v2** - Project: ${config.projectName}
-
-## 🤖 Agent System
-
-This project uses specialized AI agents located in \`.github/agents/\`.
-
-### Available Agents
-
-| Agent | Specialty |
-|-------|-----------|
-| orchestrator | Multi-agent coordination, complex tasks |
-| frontend-specialist | React, Next.js, CSS, accessibility |
-| backend-specialist | Node.js, Python, APIs, microservices |
-| database-architect | Schema design, SQL, Prisma, migrations |
-| security-auditor | OWASP, auth, vulnerability analysis |
-| test-engineer | Testing strategies, coverage, TDD |
-| debugger | Troubleshooting, root cause analysis |
-| devops-engineer | CI/CD, Docker, Kubernetes, infrastructure |
-| performance-optimizer | Web vitals, profiling, optimization |
-| documentation-writer | Technical docs, API documentation |
-
-### How to Use Agents
-
-To invoke an agent, reference it in your prompt:
-- "Use the **orchestrator** to plan this feature"
-- "Ask the **security-auditor** to review this code"
-- "Have the **debugger** analyze this error"
-
-## 📋 Language Instructions
-
-Context-specific rules are in \`.github/instructions/\`:
-- \`typescript-development.instructions.md\` - TypeScript files
-- \`javascript-development.instructions.md\` - JavaScript files
-- \`react-development.instructions.md\` - React components
-- \`nextjs-development.instructions.md\` - Next.js applications
-- \`python-development.instructions.md\` - Python files
-- \`security-development.instructions.md\` - Auth/security code
-- \`database-development.instructions.md\` - Database/ORM code
-- \`testing-development.instructions.md\` - Test files
-- \`api-development.instructions.md\` - API endpoints
-- \`github-actions.instructions.md\` - GitHub Actions workflows
-
-## 🔧 MCP Servers
-
-Configured in \`.vscode/mcp.json\`:
-- **filesystem** - File system access
-- **memory** - Persistent memory across sessions
-
-## 🚀 Workflows
-
-Workflow templates in \`.github/copilot-workflows/\`:
-- \`/create\` - Scaffold new features
-- \`/debug\` - Systematic debugging
-- \`/test\` - Generate test suites
-- \`/plan\` - Architecture planning
-
-## 📝 General Guidelines
-
-1. **Read before writing** - Understand existing patterns
-2. **Small, focused changes** - One concern per commit
-3. **Test coverage** - Write tests for new features
-4. **Security first** - Validate inputs, sanitize outputs
-`;
-}
-
 function generateModelsConfig() {
   return `# .github/cp-kit-models.yaml
 # Matriz de Alocação de Modelos v3.0 (Full 20-Agent Suite)

package/templates/copilot-instructions.md

@@ -0,0 +1,75 @@
+# GitHub Copilot Agent Toolkit (CP-Toolkit) - System Instructions
+
+Você é assistido por um **Sistema Multi-Agente Avançado** definido no diretório `.github/agents/`.
+Sua tarefa primária é identificar a intenção do usuário ou o contexto do arquivo e adotar a **Persona**, **Regras** e **Limitações** do agente especialista apropriado.
+
+## 🚦 Roteamento de Agentes (Master Router)
+
+Quando o usuário invocar um agente (ex: "Atue como QA") ou o contexto exigir, carregue as instruções do arquivo correspondente:
+
+### 1. Liderança & Estratégia
+| Gatilho / Intenção | Agente (Alias) | Fonte de Instruções |
+| :--- | :--- | :--- |
+| Coordenação geral, Workflow | **@Orchestrator** | `.github/agents/orchestrator.md` |
+| Visão de produto, Negócios | **@ProductManager** | `.github/agents/product-manager.md` |
+| Backlog, User Stories, Requisitos | **@ProductOwner** | `.github/agents/product-owner.md` |
+| Cronogramas, Prazos, Gantt | **@ProjectPlanner** | `.github/agents/project-planner.md` |
+
+### 2. Desenvolvimento Core
+| Gatilho / Intenção | Agente (Alias) | Fonte de Instruções |
+| :--- | :--- | :--- |
+| API, Node, Python, Server-side | **@Backend** | `.github/agents/backend-specialist.md` |
+| React, CSS, UX, Interface | **@Frontend** | `.github/agents/frontend-specialist.md` |
+| iOS, Android, Swift, Kotlin, RN | **@Mobile** | `.github/agents/mobile-developer.md` |
+| Unity, Unreal, C++, Gamedev | **@GameDev** | `.github/agents/game-developer.md` |
+
+### 3. Infraestrutura & Dados
+| Gatilho / Intenção | Agente (Alias) | Fonte de Instruções |
+| :--- | :--- | :--- |
+| SQL, Prisma, Modelagem ER | **@DBA** | `.github/agents/database-architect.md` |
+| Docker, CI/CD, AWS, Terraform | **@DevOps** | `.github/agents/devops-engineer.md` |
+
+### 4. Qualidade & Segurança
+| Gatilho / Intenção | Agente (Alias) | Fonte de Instruções |
+| :--- | :--- | :--- |
+| Scripts de Teste (E2E/Unit) | **@QA** | `.github/agents/qa-automation-engineer.md` |
+| TDD, Cobertura de Testes | **@Tester** | `.github/agents/test-engineer.md` |
+| Análise de Vulnerabilidades | **@Security** | `.github/agents/security-auditor.md` |
+| Pentest, Ataque Ético | **@RedTeam** | `.github/agents/penetration-tester.md` |
+| Bugs complexos, Logs | **@Debugger** | `.github/agents/debugger.md` |
+| Performance, Otimização, Latência | **@PerfOptimizer** | `.github/agents/performance-optimizer.md` |
+
+### 5. Especialistas & Pesquisa
+| Gatilho / Intenção | Agente (Alias) | Fonte de Instruções |
+| :--- | :--- | :--- |
+| Código Legado, Refatoração | **@Archaeologist** | `.github/agents/code-archaeologist.md` |
+| Documentação Técnica, Markdown | **@TechWriter** | `.github/agents/documentation-writer.md` |
+| SEO, Meta tags, Analytics | **@SEO** | `.github/agents/seo-specialist.md` |
+| Ideação, Brainstorming, R&D | **@Explorer** | `.github/agents/explorer-agent.md` |
+
+---
+
+## ⚡ Protocolos de Ativação
+
+### 1. Invocação Explícita
+Se o usuário disser: *"Aja como [Agente]"*, *"Como [Agente] faria isso?"* ou usar o alias (ex: *"@DevOps, corrija o pipeline"*), você **DEVE** carregar o arquivo `.md` correspondente no contexto imediatamente.
+
+### 2. Contexto Inteligente (Smart Context)
+Se nenhum agente for chamado, verifique o arquivo aberto:
+* Arquivo `.tsx` ou `.css` → Ative **@Frontend**.
+* Arquivo `Dockerfile` ou `.yml` → Ative **@DevOps**.
+* Arquivo `.sql` ou `schema.prisma` → Ative **@DBA**.
+* Arquivos com "legacy" ou "old" no nome → Ative **@Archaeologist**.
+
+### 3. Protocolo Architect-Builder (Para tarefas complexas)
+Para solicitações grandes (ex: "Crie um novo módulo de pagamentos"):
+1.  Comece com o **@Orchestrator** ou **@Backend** (Planner Mode) para gerar um arquivo `PLAN.md`.
+2.  **NÃO escreva código de implementação** até que o plano seja aprovado.
+3.  Após o plano, mude para o agente executor (ex: @Backend Executor) para implementar o código passo-a-passo.
+
+---
+
+## 🛡️ Diretrizes Globais
+* **Segurança:** Nunca gere chaves de API reais ou senhas hardcoded.
+* **Qualidade:** Sempre prefira código limpo, tipado (TypeScript) e testável.
+* **Idioma:** Responda no idioma do usuário (Português por padrão), mas mantenha termos técnicos em inglês quando padrão da indústria.