mcp-adk

elliemci/mcp-adk

3.2

If you are the rightful owner of mcp-adk and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to dayong@mcphub.com.

The Model Context Protocol (MCP) server is a versatile tool designed to facilitate communication between agents and servers, enabling efficient data management and task execution.

Tools
5
Resources
0
Prompts
0

ADK Agent with MCP Server

ADK Agent connected to Local and Remote MCP Servers

Connect to Notion MCP Server

To connect local ADK Agent, remote_mcp/agent.py to Notion MCP Server ensure the virtual environment is active and spin the adk web from the root directory of the mcp-adk project.

Notion MCP can be used to create documentation, search and find answers from workspace content, manage task, generate code snipets from task description, build reports, create release notes, project updates and performance reports from multiple sources etc.

Running local MCP Server

ADK Agent connected to Local Servers with SQL Tools

  1. Set Up the Local MCP Server local_mcp/server.py

  2. Create the data base local_mco/create_db.py

  3. Set up the SQL Tools: list_db_tables, get_table_schema, query_db_table, insert_data, delete_data; run the server from local_mcp directory

  4. Set the local ADK Agent local_mcp/agent.py, and connect it to the local MCP Server; spin the adk web from the porject root dir

The local ADK agent is configured to automatically start the MCP server when it initalizes its MCP toolset.

To run the agent in active virutal environment python3 local_mcp/agent

This will

  • Start the agent.py script.
  • The agent, upon initializing the MCPToolset, will execute the python3 local_mcp/server.py command.
  • The server.py (MCP server) will start and listen for tool calls from the agent via stdio.
  • The agent will then be ready to process instructions typically provided in a client application or test environment that uses this agent.
  • log output from both the agent and the MCP server in local_mcp/mcp_server_activity.log, and potentially to the console if uncommented the stream handler in server.py.