DAIV : Development AI Assistant

Development teams spend countless hours on repetitive tasks: responding to code reviews, fixing broken builds, and manually implementing feature requests. DAIV changes this by bringing AI automation directly to your existing Git workflow.

DAIV is an open-source AI assistant that automates development tasks using intelligent agents. It integrates with your repositories to handle routine work, letting you focus on creative problem-solving.

🎯 What DAIV Does

DAIV automates four key development activities:

• 🚀 Issue Resolution - Converts issue descriptions into working code automatically
• 🔄 Code Review Response - Addresses reviewer feedback and updates pull requests
• 🔍 Pipeline Repair - Monitors CI/CD failures and applies fixes automatically
• 💬 Codebase Questions - Answers questions about your code using built-in search

---

🛠️ How DAIV Works

DAIV integrates directly with Git platforms using webhooks. No separate interface required. You continue using your existing workflow while DAIV handles automation in the background.

Workflow Overview

graph TD
    A["👤 Developer"] --> B["📝 Creates Issue"]
    B --> C["🤖 DAIV Agent"]
    C --> D["📋 Generates Plan"]
    D --> E["💬 Posts Plan Comment"]

    A --> F["✅ Approves Plan"]
    F --> C
    C --> G["🔨 Executes Changes"]
    G --> H["📤 Creates Pull Request"]

    I["🚦 CI/CD Pipeline"] --> J["❌ Fails"]
    J --> C
    C --> K["🔍 Analyzes Logs"]
    K --> L["🛠️ Applies Fixes"]

    M["👥 Code Reviewer"] --> N["💬 Requests Changes"]
    N --> C
    C --> O["🔄 Updates Code"]
    O --> H

    P["📤 Code Push"] --> C
    C --> Q["🗂️ Re-indexes Codebase"]

Supported Events and Actions

DAIV monitors these repository events and responds automatically:

Repository Event	DAIV Response	Example
✨ Issue created	Generates implementation plan	"Add user authentication" → detailed technical plan
📝 Issue updated	Updates plan if title/description changed	Issue scope changes → revised plan
💬 Comment on issue	Executes approved plan	Developer approves → creates pull request with code
💬 Comment on merge request	Addresses review feedback or answers questions	"Fix the bug in line 45" → updates code
🚦 Pipeline failure	Analyzes logs and applies fixes	Test failures → code fixes + updated tests
📤 Repository push	Updates codebase index for future queries	New commits → refreshed code understanding

---

🔌 Supported Platforms

Currently supported:

• GitLab - Full feature support

Coming soon:

• GitHub - Planned for future release

Contributing

Want GitHub support sooner? Contributions are welcome!

---

⏭️ Get Started

Step 1: Installation

• Set up DAIV - Install and configure DAIV in your environment
• Connect your repository - Link DAIV to your GitLab project

Step 2: Try DAIV

• Create a test issue - Write "Add a hello world function" and watch DAIV generate a plan
• Approve the plan - Comment "Go ahead" to see DAIV create a pull request with working code

Step 3: Customize

• Configure agents - Adjust how DAIV's AI agents behave for your codebase
• Advanced settings - Fine-tune DAIV for your team's needs

Step 4: Learn More

• Join discussions - Ask questions and share experiences
• Report bugs - Help improve DAIV

---

💡 Quick Example

Here's what a typical DAIV interaction looks like:

1. Developer creates issue: "Add rate limiting to the API endpoints"
2. DAIV generates plan: Analyzes codebase and posts detailed implementation steps
3. Developer approves: Comments "Looks good, go ahead"
4. DAIV implements: Creates pull request with rate limiting code, tests, and documentation
5. Reviewer requests changes: "Use Redis instead of in-memory storage"
6. DAIV updates code: Automatically modifies the implementation to use Redis

Result: Feature implemented in minutes, not hours.