DevPilot

![# 🚀 DevPilot Banner](.github_resource/github-header-banner.jpeg "DevPilot Banner")

# 🚀 DevPilot
> **Teaching machines to build machines.**

DevPilot is an autonomous AI-powered software development pipeline that transforms ideas into structured software systems using multi-agent collaboration.

---

## 🧠 What is DevPilot?

DevPilot simulates a real software engineering team using AI agents:

- 🧑‍💼 Planner → Defines project scope
- 📄 Requirements Engineer → Writes specifications
- 🏗️ Software Architect → Designs system architecture
- 👨‍💻 Engineers (coming next) → Generate real code

Give it a simple idea → DevPilot turns it into:

- Project Plan
- Software Requirements
- Technical Architecture
- (Next) Full working application

---

## ⚙️ How It Works

```text
Project Idea (YAML)
        ↓
Planning Agent
        ↓
Requirements Agent
        ↓
Architecture Agent
        ↓
🚧 Code Generation (Phase 2)
        ↓
🚧 Self-Healing Execution (Phase 3)
```

---

## 📂 Project Structure

```text
my-ai-company-v2/
├── app/
│   ├── pipeline/
│   │   ├── flow.py              # main orchestrator
│   │   ├── state.py             # RunState dataclass / schema
│   │   ├── gates.py             # validation rules
│   │   ├── retries.py           # retry policies
│   │   └── events.py            # streaming/log events
│   │
│   ├── agents/
│   │   ├── planner.py
│   │   ├── requirements.py
│   │   ├── architect.py
│   │   ├── backend_dev.py
│   │   ├── backend_fixer.py
│   │   ├── frontend_dev.py
│   │   ├── frontend_fixer.py
│   │   ├── qa.py
│   │   ├── devops.py
│   │   └── reviewer.py
│   │
│   ├── executors/
│   │   ├── backend_executor.py
│   │   ├── frontend_executor.py
│   │   ├── test_executor.py
│   │   └── docker_runner.py
│   │
│   ├── parsers/
│   │   ├── api_contract.py
│   │   ├── token_usage.py
│   │   └── artifact_manifest.py
│   │
│   ├── storage/
│   │   ├── runs.db              # SQLite
│   │   ├── cost_tracker.py
│   │   └── memory_store.py
│   │
│   └── ui/
│       ├── server.py
│       └── templates/index.html
│
├── projects/
│   ├── bakery.yaml
│   └── school-tracker.yaml
│
├── outputs/
│   └── <run-id>/
│
├── Dockerfile.sandbox
├── docker-compose.yml
├── requirements.txt
└── run.py
```

---

## 🚀 Features

- ✅ Multi-agent AI system (CrewAI)
- ✅ YAML-driven project definitions
- ✅ Automated:
  - Planning
  - Requirements generation
  - Architecture design

- 🔜 Backend code generation (Flask)
- 🔜 Frontend generation (React)
- 🔜 Test generation
- 🔜 Self-healing (auto-debugging system)

---

## 🧪 Example Run

```bash
uv run python run.py projects/demo-flask.yaml
```

Output:

```text
STATUS -> completed_phase_1
```

Generated:

- ✅ Project Plan
- ✅ Requirements Document
- ✅ Technical Architecture

---

## 🔐 Setup

### 1. Clone the repo

```bash
git clone https://github.com/your-username/DevPilot.git
cd DevPilot
```

---

### 2. Create virtual environment

```bash
uv venv
source .venv/bin/activate
```

---

### 3. Install dependencies

```bash
uv pip install -r requirements.txt
```

---

### 4. Configure environment

Create `.env` file:

```env
OPENROUTER_API_KEY=your_api_key_here
```

---

## 🧱 Roadmap

### Phase 1 ✅ (Completed)

- Planning agent
- Requirements agent
- Architecture agent

### Phase 2 🚧 (In Progress)

- Backend code generation (Flask)
- Frontend generation (React)
- Test generation

### Phase 3 🚧

- Run generated code
- Detect failures
- Auto-fix bugs (self-healing AI)

---

## 🎯 Why DevPilot?

Modern software development is slow, repetitive, and human-limited.

DevPilot explores a different future:

> AI systems that design, build, and improve software autonomously.

This project demonstrates:

- Multi-agent system design
- AI-driven development workflows
- Automated system architecture generation
- Foundation for autonomous engineering platforms

---

## 👨‍💻 Author

**Helitha Praveen** - Backend Engineer 🇱🇰

---

## ⭐ Vision

DevPilot is not just a project.

It’s the first step toward:

> Fully autonomous AI software companies.