This project provides Bedrock Engineer, a universal AI agent application for software development tasks. It empowers users to create, edit files, execute commands, search the web, leverage knowledge bases, utilize multi-agents, and generate images, all through an interactive UI powered by Amazon Bedrock.

How It Works

Bedrock Engineer acts as an autonomous software development agent, integrating with Amazon Bedrock's generative AI models. It offers a customizable agent architecture where users can define agent roles, system prompts, and available tools. The application supports a wide array of tools, including file system operations, web search via Tavily API, image generation and recognition through Bedrock, and secure command execution. This modular tool integration allows for flexible agent customization for diverse use cases beyond traditional development.

Quick Start & Requirements

• Installation: Download the latest macOS release (DMG) or build from source.
• Prerequisites: Node.js (for building from source), AWS credentials configured for Amazon Bedrock, Tavily API key.
• Setup: For macOS, download, drag to Applications, and configure AWS credentials. May require allowing apps from "anywhere" in Security & Privacy. Building from source requires npm install and npm run build:<os>.
• Links: Demo

Highlighted Details

• Supports multiple Amazon Bedrock models (Nova, Claude, Llama) and image generation models (Stability.ai, Titan).
• Features an Agent Directory for discovering and sharing custom agents.
• Includes specialized tools for website generation (React, Vue, Svelte, Vanilla.js) and AWS Step Functions/architecture diagram generation.
• Offers Agentic-RAG for referencing design systems and codebases via Knowledge Base integration.

Maintenance & Community

The project is part of aws-samples, indicating official AWS backing. Community contributions are encouraged via pull requests and GitHub issues for agent sharing and feature development.

Licensing & Compatibility

• License: MIT-0 License.
• Compatibility: Permissive license allows for commercial use and integration with closed-source projects.

Limitations & Caveats

The application is optimized for macOS, with Windows and Linux support requiring building from source and potential troubleshooting. Security measures restrict command execution to configured commands, and users may need to adjust macOS security settings to run the application.