Transform academic trading research into executable QuantConnect algorithms using AI.

QuantCoder is a command-line tool that converts research papers into production-ready QuantConnect trading algorithms using natural language processing and large language models. Based on a dual-agent cognitive architecture, it extracts trading signals, risk management rules, and generates tested Python code.

As of November 2025, it is under refactoring with readiness expected in February 2026.

• 📄 PDF Processing: Extract trading strategies from academic papers
• 🔍 CrossRef Integration: Search and download financial research articles
• 🤖 AI-Powered Code Generation: Uses GPT-4o to generate QuantConnect algorithms
• ✅ Syntax Validation: Automatic code validation and refinement
• 🎯 Dual-Agent Architecture: Separates strategy extraction from code generation
• 📊 Rich Terminal UI: Beautiful, interactive command-line interface

• Python 3.9 or later
• OpenAI API key

# Clone the repository
git clone https://github.com/SL-Mar/quantcoder-legacy.git
cd quantcoder-legacy

# Create and activate virtual environment
python -m venv .venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate

# Install QuantCoder
pip install -e .

# Download required NLP model
python -m spacy download en_core_web_sm

# Set your OpenAI API key
echo "OPENAI_API_KEY=your-api-key-here" > .env

Launch the interactive CLI:

quantcli interactive

Or:

python -m quantcli.cli interactive

Search for research articles:

quantcli search "momentum trading strategies" --num 5

List previously searched articles:

quantcli list

Download an article:

quantcli download 1

Process a PDF to generate algorithm:

quantcli process path/to/research-paper.pdf

1. Search for trading research:

   quantcli search "mean reversion high frequency" --num 3

2. Download an interesting paper:

   quantcli download 1

3. Generate QuantConnect algorithm:

   quantcli process downloads/paper.pdf

4. Review generated code in generated_code/ directory

5. Copy to QuantConnect and backtest

QuantCoder uses a dual-agent system:

1. Extraction Agent: Analyzes PDF, identifies trading signals and risk management rules
2. Generation Agent: Converts extracted information into QuantConnect Python code
3. Validation Layer: Checks syntax and refines code using AST analysis

✅ Migrated to OpenAI SDK 1.x+ - Modern API with better error handling ✅ LLMClient abstraction layer - Easily swap LLM providers ✅ Token usage tracking - Monitor API costs ✅ Test infrastructure - pytest with coverage reporting ✅ Improved logging - Structured logs for debugging ✅ Type hints - Better code quality with mypy support

See CHANGELOG.md for full details.

Run the test suite:

pytest

With coverage:

pytest --cov=quantcli --cov-report=html

• ✅ 10K+ LinkedIn impressions on first algorithm generated
• ✅ 79 GitHub stars from quantitative trading community
• ✅ 21 forks actively used by traders worldwide

Original case study: "Outperforming the Market (1000% in 10 years)"

Create a .env file in the project root:

OPENAI_API_KEY=your-openai-api-key

Optional configuration:

# Change default model (default: gpt-4o-2024-11-20)
OPENAI_MODEL=gpt-4-turbo-preview

Contributions welcome! Please:

1. Fork the repository
2. Create a feature branch (git checkout -b feature/amazing-feature)
3. Commit your changes (git commit -m 'Add amazing feature')
4. Push to the branch (git push origin feature/amazing-feature)
5. Open a Pull Request

This project is licensed under the MIT License - see the LICENSE file for details.

• Inspired by "Dual Agent Chatbots and Expert Systems Design"
• Built for the QuantConnect algorithmic trading platform
• Powered by OpenAI GPT-4

Author: SL-MAR Email: smr.laignel@gmail.com GitHub: @SL-Mar

⭐ If QuantCoder helps your trading research, give it a star! ⭐