Package Exports
- create-sdlc-agents
- create-sdlc-agents/src/index.js
This package does not declare an exports field, so the exports above have been automatically detected and optimized by JSPM instead. If any package subpath is missing, it is recommended to post an issue to the original package (create-sdlc-agents) to support the "exports" field. If that is not possible, create a JSPM override to customize the exports field for this package.
Readme
🤖 create-sdlc-agents
18 specialized AI agents for the entire Software Development Lifecycle.
Supports both Claude Code and GitHub Copilot — plug-and-play into any project.
🚀 Quick Install
# Interactive setup wizard (recommended)
npx create-sdlc-agents
# Non-interactive with flags
npx create-sdlc-agents --platform both --type fullstack --no-interactive
# Backend API project (Claude only)
npx create-sdlc-agents --platform claude --type backend-api --no-interactive✨ Features
- 🧠 18 specialized agents — each with dedicated prompts, workflows, and guardrails
- 🔄 14 core protocols — Reflection, Debate, HITL (L1-L4), Guardrails, Error Recovery, and more
- 📚 27 skill libraries — reusable capability modules for testing, security, CI/CD, etc.
- 🔍 Auto-detect 11 languages — TypeScript, Python, Go, Java, PHP, Ruby, Rust, C#, Dart, Elixir, Swift
- 🏗️ Multi-service & monorepo support — auto-discovers sub-projects and generates
services.yml - 🔐 Dynamic MCP generation — database-aware server configs with credential placeholders
- 📋 Post-scaffold verify — built-in agent prompt to validate the entire config
- 🖥️ Cross-platform — Windows, macOS, Linux — works everywhere Node.js runs
⚡ Quick Start (5 minutes)
# 1. Scaffold agents into your project
cd my-existing-project
npx create-sdlc-agents
# 2. Configure your tech stack (auto-detected!)
# Edit: .claude/config/project.config.yml
# 3. Describe your project for AI context
# Edit: .claude/memory/project-context.md
# 4. Start using agents!
@setup-assistant Analyze my project and verify config
@implementer Create a CRUD service for the products module
@qa-tester Write unit tests for ProductService
@code-reviewer Review the products module🏗️ Architecture
User Request
│
┌────────▼────────┐
│ @router │ Classify & orchestrate
└────────┬────────┘
│
┌─────────────┼─────────────┐
│ Simple │ Complex │
▼ ▼ │
Single Agent @planner │
(Decompose) │
│ │
┌────────▼────────┐ │
│ PIPELINE │ │
│ │ │
│ Design: @architect, @db-designer, @api-designer │
│ Build: @implementer, @frontend-dev │
│ Test: @qa-tester, @code-reviewer, @security-aud │
│ Ship: @devops, @tech-writer │
└─────────────────┘🤖 18 Agents
| Agent | Role | Use When |
|---|---|---|
@router |
Classify & route requests | Automatic |
@planner |
Decompose complex tasks | Large features |
@product-manager |
PRD, user stories, acceptance criteria | New features |
@architect |
System design & architecture decisions | New modules |
@db-designer |
Database schema & migrations | Schema changes |
@api-designer |
API endpoints, DTOs, OpenAPI specs | New APIs |
@implementer |
Backend code implementation | Business logic |
@frontend-dev |
Frontend UI/UX (React, Vue, etc.) | UI components |
@qa-tester |
Unit / Integration / E2E testing | Test coverage |
@code-reviewer |
Code review with Debate Protocol | Before merge |
@security-auditor |
OWASP Top 10 security audit | Auth, payments |
@critic |
Devil's advocate, challenge assumptions | Risky decisions |
@debug-detective |
Root cause analysis & debugging | Bug reports |
@performance-optimizer |
Profiling & optimization | Slow endpoints |
@devops |
CI/CD, Docker, deployment | Infrastructure |
@tech-writer |
Documentation & changelogs | README, API docs |
@ticket-manager |
Issue triage & SLA tracking | Bug/feature tickets |
@setup-assistant |
Project setup & onboarding | New projects, config |
📦 Supported Project Types
| Type | Description | Agents |
|---|---|---|
fullstack |
Full-stack web app (default) | 18 agents |
backend-api |
REST/GraphQL API | 14 agents |
frontend-spa |
React/Vue/Angular SPA | 9 agents |
mobile-app |
React Native / Flutter | 14 agents |
cli-library |
CLI tool or npm package | 10 agents |
batch-job |
Cron jobs, workers, queues | 11 agents |
data-pipeline |
ETL, ML pipelines (Python) | 12 agents |
custom |
Pick your own agent groups | Variable |
🔍 Auto-Detection (11 Languages)
The CLI automatically detects your project's tech stack:
| Language | Detected From | Frameworks |
|---|---|---|
| TypeScript / JavaScript | package.json |
NestJS, Express, Fastify, Next.js, React, Vue, Angular, Svelte |
| Python | requirements.txt, pyproject.toml |
FastAPI, Django, Flask |
| Go | go.mod |
Gin, Fiber, Echo |
| Java / Kotlin | pom.xml, build.gradle |
Spring |
| PHP | composer.json |
Laravel, Symfony, Slim, CakePHP |
| Ruby | Gemfile |
Rails, Sinatra |
| Rust | Cargo.toml |
Actix, Axum, Rocket |
| C# / .NET | *.csproj, *.sln |
ASP.NET, Entity Framework Core |
| Dart / Flutter | pubspec.yaml |
Flutter |
| Elixir | mix.exs |
Phoenix |
| Swift | Package.swift |
— |
Also detects: ORM (Prisma, Eloquent, ActiveRecord, Doctrine, EF Core), Auth (JWT, Passport, NextAuth), Database (from docker-compose.yml or .env).
🏢 Multi-Service & Monorepo
For microservice architectures with mixed languages:
my-platform/
├── api-laravel/ ← composer.json → PHP + Laravel
├── api-go-gateway/ ← go.mod → Go + Gin
├── web-admin/ ← package.json → TypeScript + React
├── batch-etl/ ← requirements.txt → Python
├── api-rust-perf/ ← Cargo.toml → Rust + Actix
└── api-dotnet/ ← MyApi.csproj → C# + .NETcd my-platform
npx create-sdlc-agents
# → Auto-detects all 6 services
# → Generates services.yml with per-service config
# → Creates cross-service dependency map🗂️ What Gets Scaffolded
your-project/
├── .claude/ ← Claude Code agent system
│ ├── .mcp.json ← MCP servers (edit credentials here)
│ ├── CLAUDE.md ← Global rules
│ ├── agents/ (18 files) ← Agent definitions
│ ├── core/ (14 files) ← Core protocols
│ ├── skills/ (27 folders) ← Skill library
│ ├── config/
│ │ ├── project.config.yml ← Tech stack config (auto-filled)
│ │ └── services.yml ← Service registry (multi-service only)
│ └── memory/
│ ├── project-context.md ← Project description (YOU EDIT THIS)
│ ├── lessons-learned.md ← Auto-updated bug patterns
│ └── inbox/ ← Drop docs here for ingestion
│
├── .github/ ← GitHub Copilot agent system
│ ├── agents/ (18 .agent.md) ← Same agents, Copilot format
│ ├── core/, skills/, config/ ← Full parity with Claude
│ └── ...
│
├── .vscode/
│ └── mcp.json ← MCP config for VS Code / Copilot
│
└── AGENTS_README.md ← Quick start guide🔄 Cross-Platform Parity
Both systems share the same 18 agents, 14 protocols, and 27 skills — only file format differs:
| Component | Claude Code | GitHub Copilot |
|---|---|---|
| Rules file | CLAUDE.md |
copilot-instructions.md |
| Agent format | agent-name.md |
agent-name.agent.md |
| Invocation | @agent in CLI |
@agent in VS Code Chat |
| MCP config | .claude/.mcp.json |
.vscode/mcp.json |
📖 Documentation
| Document | Description |
|---|---|
| SETUP_GUIDE.md | Step-by-step setup for 9 project scenarios |
| NPX_GUIDE.md | How the CLI works internally |
| Portability Guide | Migrate between Claude / Copilot / Cursor / Cline |
🔒 Security
- MCP files are generated with
<YOUR_TOKEN>placeholders — no credentials leak - Add to
.gitignore:.claude/.mcp.jsonand.vscode/mcp.json - Built-in guardrails block dangerous commands (
DROP TABLE,rm -rf, etc.)
📊 Stats
| Metric | Claude | GitHub | Total |
|---|---|---|---|
| Files | 85 | 109 | 194 |
| Total size | 507KB | 704KB | 1.2MB |
| Agent avg depth | 9.1KB | 9.2KB | — |
| Protocol avg depth | 10.9KB | 10.1KB | — |
🤝 Contributing
- Fork the repository
- Create your feature branch (
git checkout -b feature/amazing-feature) - Run tests:
npm test(195 assertions across 18 test suites) - Commit your changes (
git commit -m 'Add amazing feature') - Push and open a Pull Request
📄 License
MIT — see LICENSE for details.
Enterprise SDLC Agent System v5.0 — A+ Grade (9.55-9.58/10)
© 2026 Doan Hai Duy — Built with Claude Opus 4.6