Documentation-Driven Development v2 For Agentic AI puts your project's documentation at the center of your AI workflow.
Write markdown or html specs → Agentic co-writing of context files → Vibe-coding with a targeted context → Commit doc for archive between steps
📋 How It Works
After running DDD: Initialize Project, you get:
.ai_context/
├── README.md ← project overview (permanent)
├── CONTRACT.md ← agent interaction rules (permanent)
├── CONTEXT.md ← project identity: stack, conventions, team (permanent — never reset)
├── context.json ← current task metadata
├── vision.md ← product vision and epic goals (semi-permanent — reset on New Vision)
├── history.json ← append-only log of all visions and tasks
├── steps/ ← roadmap phases for current vision (reset on New Vision)
│ ├── phase1-core.md
│ └── phase2-growth.md
├── skills/ ← reusable agent knowledge (permanent-* kept)
│ ├── permanent-dev-stack.md
│ └── permanent-architecture.md
└── tasks/ ← current sprint work (reset on New Tasks)
├── done/ ← AI execution reports
├── specification/ ← functional specs (permanent-* kept)
└── technical/ ← technical decisions (permanent-* kept)
Define WHY and WHAT — your project's purpose, goals, and architectural decisions.
Break down HOW — milestones, sprints, feature phases. Each step adds to AI context.
Describe your TECH CHOICES — frameworks, libraries, patterns.
Document WHO knows WHAT — team expertise, conventions, gotchas.
🚀 Key Features
✅ Initialize Project — One-click setup with auto-detected stack templates
✅ Create Vision Docs — Document project goals and architecture
✅ Create Task Lists — Break down development into actionable steps
✅ Manage Skills — Record team expertise and capabilities
✅ Tree Explorer — Navigate all DDD files in the sidebar
📖 Learn More
For full documentation, see HOWTO.md
MIT License — Copyright © 2024 Documentation First