_______ ___ _______.____ ____ ___ _______ _______ .__ __. .___________.
| ____| / \ / |\ \ / / / \ / _____|| ____|| \ | | | |
| |__ / ^ \ | (----` \ \/ / / ^ \ | | __ | |__ | \| | `---| |----`
| __| / /_\ \ \ \ \_ _/ / /_\ \ | | |_ | | __| | . ` | | |
| |____ / _____ \ .----) | | | / _____ \ | |__| | | |____ | |\ | | |
|_______/__/ \__\ |_______/ |__| /__/ \__\ \______| |_______||__| \__| |__|
EasyAgent is a beginner-friendly intelligent agent project built on LangChain and LangGraph, integrating various practical MCP tools and services to help users complete various basic tasks.
- 🤖 Intelligent workflow based on LangChain and LangGraph
- 🔧 Integration of multiple MCP tools and services:
- Web search service
- AKShare A-share data
- Yahoo Finance financial data
- ArXiv research paper library
- Amap geographic service
- ...
- 📊 Financial data analysis support
- 🔍 Intelligent search capabilities
- 🎨 Simple and intuitive web interface
- Python 3.12
- Node.js (optional, for running certain MCP services)
- Clone the repository:
git clone https://github.com/Yangyf1125/EasyAgent.git
cd EasyAgent
- Install Python dependencies:
pip install -r requirements.txt
Configure LLM settings in the 🐋Deepseek page of the web interface or in config/llm_config.json
. Currently only Deepseek is supported:
api_key
: Your Deepseek API key
Configure various MCP tools and services in config/mcp_config.json
:
{
"server1": {
"type": "sse",
"url": "http://localhost:8000/sse"
}
}
// Other MCP service configurations...
Currently using Modelscope's remote MCP server, please modify as needed
Provides both terminal and web interfaces.
Start the terminal interface:
python main.py
Provides a simple web interface that can be deployed locally or accessed directly through the web service deployed on Streamlit Cloud.
- Start the web service:
streamlit run Homepage.py
- Access the web interface at the URL provided in your terminal(e.g.,
http://localhost:8501
)
EasyAgent/
├── src/ # Source code directory
│ ├── config/ # Configuration files
│ ├── workflow/ # Workflow related code
│ └── tool/ # Tool implementations
├── pages/ # Web page files
├── config/ # Configuration directory
├── main.py # CLI entry point
├── Homepage.py # Web interface entry point
├── requirements.txt # Python dependencies
└── README.md # Project documentation
- Please ensure all necessary API keys are properly configured before use
- Some features require internet connection
- It is recommended to run the project in a virtual environment
- Make sure all necessary dependencies are installed
This project is licensed under the MIT License. See the LICENSE file for details.
Issues and Pull Requests are welcome to help improve the project.