Adk Mcp Test
What is Adk Mcp Test
adk-mcp-test is a testing framework for the Google Agent Development Kit (ADK) and the Model Context Protocol (MCP). It includes basic test code for developing agents that can perform various tasks using these technologies.
Use cases
Use cases include developing agents for database access, investment research, and company data analysis, as well as educational purposes for understanding ADK and MCP functionalities.
How to use
To use adk-mcp-test, clone the repository and set up the necessary environment variables for database connections. Run the agents using the command uv run adk web to start the testing process.
Key features
Key features include a simple Hello World agent for basic tasks like rolling dice and Base64 encoding, a Memory Agent that interacts with a Neo4j knowledge graph, and a more advanced multi-agent setup that utilizes the Diffbot Companies+Articles dataset.
Where to use
adk-mcp-test can be used in fields such as artificial intelligence, data analysis, and software development, particularly for creating and testing intelligent agents that utilize knowledge graphs.
Clients Supporting MCP
The following are the main client software that supports the Model Context Protocol. Click the link to visit the official website for more information.
Overview
What is Adk Mcp Test
adk-mcp-test is a testing framework for the Google Agent Development Kit (ADK) and the Model Context Protocol (MCP). It includes basic test code for developing agents that can perform various tasks using these technologies.
Use cases
Use cases include developing agents for database access, investment research, and company data analysis, as well as educational purposes for understanding ADK and MCP functionalities.
How to use
To use adk-mcp-test, clone the repository and set up the necessary environment variables for database connections. Run the agents using the command uv run adk web to start the testing process.
Key features
Key features include a simple Hello World agent for basic tasks like rolling dice and Base64 encoding, a Memory Agent that interacts with a Neo4j knowledge graph, and a more advanced multi-agent setup that utilizes the Diffbot Companies+Articles dataset.
Where to use
adk-mcp-test can be used in fields such as artificial intelligence, data analysis, and software development, particularly for creating and testing intelligent agents that utilize knowledge graphs.
Clients Supporting MCP
The following are the main client software that supports the Model Context Protocol. Click the link to visit the official website for more information.
Content
Agent Development Kit (ADK) and MCP Test
This repository contains some basic test code for Agent Development Kit (ADK) and the Model Context Protocol (MCP) Servers.
Hello World Agent
Simple agent with these tools:
- Rolling dice
- Base 64 encoding
- Doodle access from a file / URL
uv run adk web
Memory Agent
Simple Agent that access MCP STDIO Tools and uses neo4j-mcp-cypher to access a memory Neo4j knowledge graph.
Add a neo4j-memory.env with Database connection information:
NEO4J_URI=bolt://localhost:7687 NEO4J_USERNAME=neo4j NEO4J_PASSWORD=your_password # OPENAPI_API_KEY=your_openapi_key # ANTHROPIC_API_KEY=your_anthropic_key
Then run the agent:
uv run adk web
More Advanced MCP + ADK Agent
Uses the Diffbot Companies+Articles dataset and the MCP Toolbox to access a Neo4j knowledge graph with company data.
Multi-Agent Setup with 3 Agents
- Database Agent
- Investment Agent
- Investor Research Agent
- Root Agent
Tools
-
Python tools for database access
-
MCP Toolbox for accessing the knowledge graph with predefined tools
Dev Tools Supporting MCP
The following are the main code editors that support the Model Context Protocol. Click the link to visit the official website for more information.
