ADK Agent Suite

A collection of specialized AI agents built with Google's Agent Development Kit (ADK), designed to perform various tasks with seamless interoperability through the Agent-to-Agent (A2A) protocol.

🤖 Featured Agents

🔌 Async MCP Connector (Main Agent)

A dynamic connector agent that can discover, connect to, and interact with any Model Context Protocol (MCP) server at runtime. Simply provide a name, connection details, and environment variables, and the agent will:

• Dynamically connect to MCP servers without prior configuration
• Handle both SSE and STDIO connection types
• Manage environment variables securely
• Expose MCP tools for interactive use

🔍 Deep Wiki Agent (Sub Agent)

Leverages DeepWiki's documentation capabilities to:

• Explore and understand public GitHub repositories
• Access structured documentation about codebases
• Answer questions about repositories using AI-powered search
• Provide context-grounded information without fabrication

📊 Reddit Scout (Sub Agent)

Monitors and analyzes Reddit for AI-related trends and discussions:

• Fetches hot threads from AI-focused subreddits
• Identifies emerging topics and discussions
• Tracks community sentiment on AI technologies
• Supports research and trend analysis

🗣️ Speaker Agent (Sub Agent)

Converts text to speech using ElevenLabs' advanced TTS technology:

• Transforms written content into natural-sounding audio
• Supports various speaking styles and voices
• Provides audio URLs for easy access and sharing
• Ideal for accessibility and content consumption on-the-go

📝 Summarizer Agent (Sub Agent)

Condenses and extracts key information from longer texts:

• Creates concise summaries of lengthy documents
• Identifies main points and critical insights
• Helps with information overload and decision-making
• Maintains the essential meaning of original content

🎭 Coordinator Agent (Main Agent)

Orchestrates multiple agents to perform complex workflows:

• Routes requests to specialized sub-agents
• Manages Reddit post discovery, summarization, and text-to-speech conversion
• Handles deep wiki searches for documentation
• Creates seamless multi-agent experiences

🔄 Agent-to-Agent (A2A) Protocol

This repository implements Google's Agent-to-Agent (A2A) protocol, enabling seamless communication between AI agents built on different frameworks:

• Standardized Communication: Uses JSON-RPC 2.0 over HTTP(S)
• Agent Discovery: Implemented via "Agent Cards"
• Flexible Interaction: Supports synchronous, streaming (SSE), and asynchronous push notifications
• Rich Data Exchange: Handles text, files, and structured JSON
• Enterprise-Ready Security: Built with security considerations in mind

🚀 Getting Started

1. Clone this repository
2. Install dependencies: pip install -r requirements.txt
3. Set up environment variables in a .env file
4. Run an agent: adk web --app=agents/[agent_directory]

📋 Requirements

• Python 3.9+
• Google ADK 1.2.1+

🤝 Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

📄 License

This project is licensed under the terms of the MIT license.