MCP-Langgraph Integration Tutorial
This tutorial demonstrates how to integrate Model Context Protocol (MCP) servers with Langgraph agents to create powerful, tool-enabled AI applications. The project showcases a data science assistant named Scout that can help users manage their data science projects using various MCP-powered tools.
Overview
The project implements a conversational AI agent that:
- Uses GPT-4.1 as the base model
- Integrates with multiple MCP servers for different functionalities
- Uses Langgraph for orchestrating the conversation flow
- Provides a streaming interface for real-time responses
Prerequisites
- Python 3.13+
- Node.js (for filesystem MCP server)
- Docker (for GitHub MCP server)
- UV package manager
- OpenAI API key
Project Structure
scout/
├── graph.py # Langgraph agent implementation
├── client.py # MCP client and streaming interface
├── client_utils.py # Utility functions
├── main.py # Entry point
└── my_mcp/ # MCP server configurations
├── config.py # Config loading and env var resolution
├── mcp_config.json # MCP server definitions
└── local_servers/ # Custom MCP server implementations
Setup
- Clone the repository:
git clone <repository-url>
cd mcp-intro
- Create and activate a virtual environment:
python -m venv .venv
source .venv/bin/activate # On Windows: .venv\Scripts\activate
- Install dependencies:
uv pip install -e .
- Set up environment variables:
Create a
.envfile with:
OPENAI_API_KEY=your_openai_api_key
MCP_FILESYSTEM_DIR=/path/to/projects/directory
MCP_GITHUB_PAT=your_github_personal_access_token
MCP Servers
This project integrates with four MCP servers:
- Dataflow Server: Custom implementation for data loading and querying
- Filesystem Server: Uses
@modelcontextprotocol/server-filesystemfor file operations - Git Server: Uses
mcp-server-gitfor local git operations - GitHub Server: Uses the official GitHub MCP server for GitHub operations
Usage
- Start the application:
python -m scout.client
- Interact with Scout by typing your questions or requests. For example:
USER: Can you help me set up a new data science project?
- Scout will use its tools to:
- Create and manage project directories
- Handle data loading and transformation
- Manage version control
- Interact with GitHub repositories
- Type 'quit' or 'exit' to end the session.
How It Works
- The
graph.pyfile defines the Langgraph agent structure:
- Sets up the system prompt and agent state
- Configures the LLM (GPT-4)
- Defines the conversation flow graph
- The
client.pyfile:
- Initializes the MCP client with multiple servers
- Handles streaming responses
- Manages the interactive session
- MCP servers provide tools for:
- File system operations
- Data manipulation
- Git operations
- GitHub interactions
Extending the Project
You can extend this project by:
- Adding new MCP servers in
my_mcp/local_servers/ - Modifying the system prompt in
graph.py - Adding new tools to the agent
- Customizing the conversation flow
Contributing
Contributions are welcome! Please feel free to submit a Pull Request.
Metadata
Release files for iflow-mcp_kenneth-liao-mcp-intro 0.1.0
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | Actions |
|---|---|---|---|---|
| iflow_mcp_kenneth_liao_mcp_intro-0.1.0.tar.gz | 8.3 kB | Without provenance | Details for iflow_mcp_kenneth_liao_mcp_intro-0.1.0.tar.gz |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| iflow_mcp_kenneth_liao_mcp_intro-0.1.0-py3-none-any.whl | Python 3 | none | any | Details for iflow_mcp_kenneth_liao_mcp_intro-0.1.0-py3-none-any.whl |
Total release size: 18.5 kB
Release files / iflow_mcp_kenneth_liao_mcp_intro-0.1.0.tar.gz
| Download URL | iflow_mcp_kenneth_liao_mcp_intro-0.1.0.tar.gz |
|---|---|
| Size | 8.3 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
bfc4c6cc16d597d000c908abb857c2a915e17c2fc03a038c6059a1e1e227fa35
|
|
BLAKE2b-256 checksum How to use checksums |
863de7f6f793c7ff6d17a1587518f9a5818e71ff35dc7316577bd5fb3734b304
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
uv/0.10.0 {"installer":{"name":"uv","version":"0.10.0","subcommand":["publish"]},"python":null,"implementation":{"name":null,"version":null},"distro":{"name":"Debian GNU/Linux","version":"13","id":"trixie","libc":null},"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":null}
|
Release files / iflow_mcp_kenneth_liao_mcp_intro-0.1.0-py3-none-any.whl
| Download URL | iflow_mcp_kenneth_liao_mcp_intro-0.1.0-py3-none-any.whl |
|---|---|
| Size | 10.2 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
3bd3fab79d37fa0e81a85178d64b4d73a71d643f476675331dd528e6a99c53c1
|
|
BLAKE2b-256 checksum How to use checksums |
ac56520cb5db02f629751f56b41e36dec2d3b768ed710130ba5908253f3132e0
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
uv/0.10.0 {"installer":{"name":"uv","version":"0.10.0","subcommand":["publish"]},"python":null,"implementation":{"name":null,"version":null},"distro":{"name":"Debian GNU/Linux","version":"13","id":"trixie","libc":null},"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":null}
|
Security for iflow-mcp_kenneth-liao-mcp-intro 0.1.0
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.