A research knowledge base for the next generation of software engineering:
AI-Agent Orchestration & Agentic Coding.
This repository is a structured knowledge base documenting research on Agentic Coding — the practice of orchestrating AI agents to build software systematically, moving beyond simple "chat-based coding" (Vibe Coding) toward reliable, spec-driven, multi-agent development.
Each topic is documented using a consistent 4-document structure:
| # | Document | Purpose |
|---|---|---|
| 0 | Overview | What it is, problems solved, design philosophy |
| 1 | Architecture & Logic | Internal architecture, data flow, core mechanisms |
| 2 | Playbook | Practical guide — setup, daily operations, cheat sheet, troubleshooting |
| 3 | Decision Guide | Comparison with alternatives, use cases, costs, migration path |
Detailed research on the Claude Code agentic coding platform and its extensibility systems.
| Topic | Overview | Architecture | Playbook | Decision Guide |
|---|---|---|---|---|
| How Claude Code Works | 📖 | 🏗️ | 🛠️ | ⚖️ |
| Claude Code Hooks | 📖 | 🏗️ | 🛠️ | ⚖️ |
| Claude Code Skills | 📖 | 🏗️ | 🛠️ | ⚖️ |
| Claude Code Plugins | 📖 | 🏗️ | 🛠️ | ⚖️ |
| Claude Code Sub-agents | 📖 | 🏗️ | 🛠️ | ⚖️ |
| Claude Code Memory | 📖 | 🏗️ | 🛠️ | ⚖️ |
Detailed research on Google's open-source AI agent for the terminal.
| Topic | Overview | Architecture | Playbook | Comparison |
|---|---|---|---|---|
| Gemini CLI | 📖 | 🏗️ | 🛠️ | ⚖️ |
Research on frameworks that add structure, memory, and orchestration to AI-assisted development.
| Topic | Overview | Architecture | Playbook | Decision Guide |
|---|---|---|---|---|
| GSD (Get Shit Done) | 📖 | 🏗️ | 🛠️ | ⚖️ |
| BMad Method | 📖 | 🏗️ | 🛠️ | ⚖️ |
| Superpowers | 📖 | 🏗️ | 🛠️ | ⚖️ |
| Topic | Link | Description |
|---|---|---|
| Scrum for Agentic Teams | 📖 Index | Adapting Scrum for AI-augmented teams — Team Structure · Glossary |
| Beads: Agent Task Management | 📖 Read | Issue tracking designed for AI agents, solving "Agent Amnesia" |
| Getting Started with Claude Code | 📖 Read | Beginner-to-advanced guide with hooks and security gates |
New to agentic coding? Start here:
- How Claude Code Works — Overview — Understand the platform
- Gemini CLI — Overview — Google's open-source alternative
- GSD — Overview — Learn structured AI development
- Superpowers — Overview — TDD + quality with AI
Want to compare platforms & frameworks? Read the Decision Guides:
- Gemini CLI vs Claude Code — Side-by-side platform comparison
- GSD Decision Guide — GSD vs BMad vs Vibe-coding
- Superpowers Decision Guide — Quality-focused comparison
- Claude Code Decision Guide — Platform comparison
Ready to use immediately? Jump to the Playbooks:
- Claude Code Playbook — Setup & daily operations
- Gemini CLI Playbook — Installation, cheat sheet & troubleshooting
- GSD Playbook — Installation to production workflow
- Hooks Playbook — Event-driven automation
agentic-coding-lab/
├── README.md # 🇬🇧 You are here
├── README.vn.md # 🇻🇳 Vietnamese version
├── en/ # 🇬🇧 English documentation
│ ├── claude-code/
│ │ ├── how-claude-code-works/ # Platform internals
│ │ ├── claude-code-hooks/ # Event-driven automation
│ │ ├── claude-code-skills/ # Custom instruction sets
│ │ ├── claude-code-plugins/ # Extensibility & marketplace
│ │ ├── claude-code-sub-agents/ # Multi-agent delegation
│ │ └── claude-code-memory/ # Persistent memory & CLAUDE.md
│ ├── gemini-cli/ # Google's Gemini CLI platform
│ ├── get-shit-done/ # GSD framework
│ ├── bmad-method/ # BMad framework
│ ├── superpowers/ # Superpowers framework
│ ├── understand-scrum/ # Scrum adaptation
│ ├── understand-beads.md # Agent task management
│ └── getting-started-with-claude-code.md
└── vn/ # 🇻🇳 Vietnamese documentation
└── (mirror structure of en/)
This is a living research repo. If you find new agentic patterns, want to add new framework research, or improve translations — feel free to open a PR.
Maintained by Astragentic Team · Research is ongoing · Last updated: 2026-03