mcp_session13

djdinnebeil/mcp_session13

3.1

If you are the rightful owner of mcp_session13 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.

This repository demonstrates an MCP server with multiple tools and two methods to connect it to a LangGraph agent.

Tools
4
Resources
0
Prompts
0

MCP Server + LangGraph Demos

This repo shows an MCP server with multiple tools and two ways to connect it to a LangGraph agent.

Tools (in server.py)

  • web_search (Tavily)
  • roll_dice
  • select_name
  • get_current_weather (WeatherAPI)

A minimal, weather-only server is also provided as weather_server.py.

Two agent styles

ScriptStyleWhy use it
prebuilt_react_demo.pyPrebuilt ReActFastest to ship; minimal code (create_react_agent)
explicit_stategraph_demo.pyManual StateGraphFull control (nodes, edges, conditions)

Requirements

  • Python 3.13+
  • TAVILY_API_KEY (Tavily)
  • WEATHERAPI_API_KEY (WeatherAPI)

Quickstart

uv sync

Create .env:

TAVILY_API_KEY=your_tavily_key
WEATHERAPI_API_KEY=your_weatherapi_key

Run the multi-tool MCP server (optional)

Agents can auto-start the server via stdio. To run it yourself:

uv run server.py

Run a LangGraph demo

python prebuilt_react_demo.py
# or
python explicit_stategraph_demo.py

Try:

What’s the weather in 07054?

Customizing

  • New tools can be added by defining @mcp.tool() functions in server.py.