owl/community_usecase/stock-analysis/README.md

📈 Stock Analysis Agent

简体中文 | English

Intelligent Stock Analysis Agent Based on 🦉OWL Framework

📖 Introduction

A stock analysis agent based on the 🦉OWL framework that provides users with comprehensive stock analysis reports, including basic stock information, technical indicators, risk metrics, and investment recommendations.

Stock Analysis Agent Architecture

• Stock Analysis Agent: Uses the RolePlaying Agent from the Camel-ai framework (same as OWL) as the main agent
• Stock Analysis Tools: Utilizes report search and SEC tools to collect company basic information, financial reports, and other data
  • Search Tool: Uses search engines like Baidu (built-in tool in the Camel-ai framework)
  • SEC Tool: Retrieves company basic information and financial statements. Note: Financial statements can be hundreds of thousands of words long, so it's recommended to summarize them before use to avoid high token costs
  • SEC Agent: Uses a ChatAgent that automatically calls the SEC Tool to retrieve company financial data and generate summary reports based on the provided stock code. Free LLM models like Zhipu's GLM-4-Flash can be used here
  • Report Write Tool: Uses a file editing tool to write complete company investment analysis reports to files

🚀 Quick Start

1. Install the OWL Framework

# Clone the GitHub repository
git clone https://github.com/camel-ai/owl.git

# Navigate to the project directory
cd owl

# If you haven't installed uv yet, install it first
pip install uv

# Create a virtual environment and install dependencies
# We support Python 3.10, 3.11, 3.12
uv venv .venv --python=3.10

# Activate the virtual environment
# For macOS/Linux
source .venv/bin/activate
# For Windows
.venv\Scripts\activate

# Install CAMEL and all its dependencies
uv pip install -e .

# Navigate to the Stock Analysis Agent directory
cd community_usecase/stock-analysis

2. Install Additional SEC Tools

# Install SEC tools
uv pip install sec-api

3. Configure Environment Variables

# Create .env file
touch .env

Add relevant API keys to the .env file (refer to the .env.example file)

# DeepSeek API (https://platform.deepseek.com/api_keys)
DEEPSEEK_API_KEY='Your_Key'
DEEPSEEK_API_BASE_URL="https://api.deepseek.com/v1"

# ZHIPU API (https://bigmodel.cn/usercenter/proj-mgmt/apikeys)
ZHIPUAI_API_KEY='Your_Key'
ZHIPUAI_API_BASE_URL="https://open.bigmodel.cn/api/paas/v4/"

# SEC-API (https://sec-api.io/profile)
SEC_API_API_KEY='Your_Key'

# AgentOps API (https://app.agentops.ai/settings/billing)
AGENTOPS_API_KEY= 'Your_Key'

Tip

The project uses DeepSeek as the main model for the Stock Analysis Agent and Zhipu's GLM-4-Flash as the model for the SEC Agent

4. Run Stock Analysis

• View run parameters

python run.py --h

usage: run.py [-h] [--company COMPANY] [--use-agentops] [--rounds ROUNDS]

Stock Analysis Agent

options:
  -h, --help         show this help message and exit
  --company COMPANY  Company name to analyze
  --use-agentops     Enable AgentOps tracking
  --rounds ROUNDS    Maximum conversation rounds

• Execute company stock investment analysis

python run.py --company Apple

• View execution results

# ./log directory
Apple_chat_history.json # Records the entire execution process, including conversation history and tool call information
# ./output directory
Apple_analysis_report.md # Output investment analysis report

• View example runs
  • Apple
    • Chat History
    • Report
  • Google
    • Chat History
    • Report
  • Alibaba
    • Chat History
    • Report

🥰 Getting Help

If you encounter issues while running the project, you can try the following methods:

1. Check the error messages in the console output
2. Submit an issue on the GitHub repository

📂 Project Structure

stock-analysis
├── agent
│   └── sec_agent.py    # SEC Agent
├── example
├── log                 # log directory
├── output              # Report output directory
├── prompts.py          # Prompt templates
├── run.py              # Main file
└── tools
    └── sec_tools.py    # SEC Tool

📝 License

This project is built on the CAMEL-AI OWL framework, which is licensed under the Apache License 2.0

🙏 Acknowledgements

• This project is built on the CAMEL-AI OWL framework
• Special thanks to the contributors of CAMEL-AI

Finding the Scaling Law of Agents: The First and the Best Multi-Agent Framework.