Awesome AI DevKit

AI Coding Configuration Ecosystem — modular scenarios that turn AI coding tools into full development teams. Runs on any platform that supports the Expert plugin standard.

🇨🇳 中文 | 🌐 English

What is this?

Awesome AI DevKit is a cross-platform scenario-based AI coding configuration repository. Each scenario provides a tailored team of AI roles, skills, workflows, and best practices for a specific domain.

This is a standard and configuration repository — scenarios, roles, skills, and workflows are modular, mix-and-match.

Works on any tool that supports the Expert plugin standard.

Key Capabilities

• Instruction Grooming — auto-polish vague user instructions into structured prompts, with intelligent defaults and project-aware inference
• AI Agent Security — prompt injection defense, sub-agent depth control, skill integrity verification, resource limits
• Quality Assurance — automated testing workflows, code quality analysis, release governance, semantic versioning
• Context Optimization — intelligent context compression and memory management across long sessions
• Multi-Role Teams — each scenario provides a domain-specific team of product, engineering, QA, and operations roles
• Scaffold Workflows — YAML-defined multi-step development protocols with built-in quality gates

Included Scenarios

Fullstack Dev Team (included)

A complete product-to-operations pipeline: 11 professional roles (PM, Architect, Backend, Frontend, Design System, QA, Code Review, Security, DevOps, SRE, Database Engineer) and 21 reusable skills (code quality, security, context, automation, and more).

More scenarios (business-ops, data-engineering, ml-native, mobile, etc.) can be added — see scenarios/ for the full list and contribution guide.

Quick Start

1. Clone

git clone https://gitcode.com/badhope/awesome-ai-devkit.git
cd awesome-ai-devkit

2. Bootstrap the Fullstack scenario

# Linux / macOS / WSL
bash scenarios/programming/fullstack/bootstrap.sh

# Windows PowerShell
powershell -File scenarios/programming/fullstack/bootstrap.ps1

This verifies prerequisites (git, node, python) and creates context/project.yaml from template.

3. Configure

Edit scenarios/programming/fullstack/context/project.yaml and fill in:

• Project name
• Tech stack (frontend, backend, database)
• Any project-specific conventions

4. Start Using

Open your AI coding tool in the project directory. Load the Fullstack scenario, then:

"Use the fullstack dev team to implement [your feature]"

Available workflows inside the Fullstack scenario:

• scaffolds/feature-development.yaml — new feature implementation
• scaffolds/refactoring.yaml — code refactoring
• scaffolds/incident-response.yaml — production incident handling

Install as Expert Plugin

Follow your AI coding tool's plugin installation guide. Generally:

cd /path/to/awesome-ai-devkit
# Refer to your tool's documentation for Expert plugin installation

Registers the Fullstack scenario and other included content as an Expert in your workspace.

Project Structure

awesome-ai-devkit/
├── scenarios/
│   ├── programming/fullstack/   # Fullstack Dev Team scenario
│   │   ├── agents/              # Role definitions
│   │   ├── skills/              # Reusable skill modules
│   │   ├── scaffolds/           # Workflow templates
│   │   ├── workflows/           # Step-by-step playbooks
│   │   ├── hooks/               # Lifecycle hooks
│   │   ├── context/             # Project config templates
│   │   ├── mcp/                 # MCP server configs
│   │   └── lib/                 # Shared utility scripts
│   └── ...                      # More scenarios (business-ops, ml, mobile, etc.)
├── mcp/config/                  # Platform-specific MCP settings
├── roles/expanded/              # Extended role definitions
├── docs/                        # Design docs
└── experts/                     # Expert plugin manifests (platform-agnostic)

Source Code

Links

• Website: https://x33834.github.io/awesome-ai-devkit/
• Releases: https://github.com/X33834/awesome-ai-devkit/releases
• Report issues: https://github.com/X33834/awesome-ai-devkit/issues

License

MIT © badhope