Skip to content

DeployBot

We’ve developed an agent that automates the software development lifecycle.

Code Video Cerebral Valley

Core trick

Agent clones a repo, discovers and runs its tests, then deploys to Cloud Run through MCP cloud connectors after user confirmation.

Tags

dev toolsagents

Description

We’ve developed an agent that automates the software development lifecycle. At this stage, we’ve focused on automating the second phase—testing, deployment, and monitoring. The agent includes MCP connectors that allow it to integrate with multiple cloud providers, including Azure, AWS, and Google Cloud
Sra1panasa/google_hackathon_deploy_bot★ 01 forksPythonCreated Nov 15, 2025 · last push Nov 15, 2025

README excerpt

# 🚀 Auto-Deploy Bot - Google Hackathon

An AI-powered deployment automation agent that streamlines cloud deployments across AWS, Google Cloud, and Azure.

## 🎯 Concept

The Auto-Deploy Bot is an intelligent agent that automates the entire deployment workflow:

1. 🤖 **Ask**: Which cloud platform? (AWS, GCP, Azure)
2. 🔗 **Request**: GitHub repository URL
3. 📥 **Clone**: Repository automatically
4. 🧪 **Test**: Discover and run all tests
5. ✅ **Confirm**: Ask user before deploying
6. 🚀 **Deploy**: To Cloud Run (with UI)
7. 🌐 **Deliver**: Live service URL

## 📁 Project Structure

```
hackathon/
├── deploy_agent/          # Main deployment automation agent
│   ├── agent.py          # Agent with deployment tools
│   └── requirements.txt  # Dependencies
│
└── my_agent/             # Sample agent for testing
    ├── agent.py          # Simple time-telling agent
    ├── test_agent.py     # Tests for the sample agent
    └── requirements.txt  # Dependencies
```

## 🚀 Quick Start

### Prerequisites

- Python 3.9+
- Docker Desktop (running)
- Google Cloud SDK (authenticated)
- Git

### 1. Install Dependencies

```bash
cd deploy_agent
pip install -r requirements.txt
```

### 2. Set Environment Variables

```bash
export GOOGLE_CLOUD_PROJECT=mbs-graphrag
export GOOGLE_CLOUD_LOCATION=europe-west2
```

### 3. Run the Deployment Agent Locally

```bash
cd deploy_agent
adk web --port 8000
```

Open browser: `http://localhost:8000`

### 4. Interact with the Agent

The agent will guide you through:

Team