thank you for this. Using my own browser helps me automate tasks on sites I 'd typically get detected using automation. Works like a charm! Hope you continue to work on the repo.
well it is an MIT licensed educational open source project but I can add a disclaimer to make it more transparent. Thanks for the comment.
DISCLAIMER: This project is purely educational. The AI-generated personas are fictional and do not represent real statements or endorsements from Warren Buffett, Ray Dalio, Cathie Wood, Peter Lynch, or Michael Burry.
As an educational endeavor, with the launch of Claude 3.7 and while waiting for $NVDA (NVIDIA Corporation) to publish results, I decided to fool around with langchain and created a multi-agent system that simulates a debate between famous investors analyzing stock investment opportunities.
You can find the github repo if you want to check it out.
Currently it is hardcoded to have Warren Buffett, Ray Dalio, Cathie Wood, Peter Lynch and Michael Burry debate about $MSFT (Microsoft) based on analysis provided to them by a team of analyst agents.
github.com/weirdapps/finagents
## Overview
FinAgents is an educational project. It creates a virtual investment committee where AI agents with personas of famous investors like Warren Buffett debate whether stocks are good investments. The system:
1. Gathers real financial data using yfinance
2. Has specialized analyst agents evaluate different aspects of the stock
3. Has investor agents with unique investment philosophies debate the merits
4. Synthesizes a comprehensive investment recommendation
## Key Features
- *Parallel Processing*: Agents run concurrently for faster analysis
- *Diverse Perspectives*: 5 famous investor personas with unique philosophies
- *Specialized Analysis*: 5 types of analyst agents focus on different aspects
- *Real Market Data*: Uses yfinance to pull actual stock information
- *Comprehensive Output*: Generates detailed reports saved as Markdown files
The system will analyze Microsoft (MSFT) by default, using all agents in parallel, and save detailed reports to the `results` directory.
## Output Files
For each analyzed stock (e.g., MSFT):
- results/MSFT_decision.md: Final synthesized investment recommendation
- results/MSFT_detailed_analysis.md: Detailed reports from all analysts and investors
## Requirements
- Python 3.8+
- Anthropic API key (for Claude 3.7)
- Internet connection (for financial data retrieval)