Energy ABM LLM Input Builder

LLM-powered interface for creating custom input scenarios for energy ABMs.

Team

• José: Backend/LLM Integration
• Julian: CSV Operations
• Toni: Frontend/UI

Quick Start

1. Environment Setup

# Clone and enter project
git clone https://github.com/macro-cosm/energy-llm-inputs.git
cd energy-llm-inputs

# Install with uv
uv sync --extra dev

# Set up environment variables
cp .env.template .env
# Edit .env and add your ANTHROPIC_API_KEY

2. Run Backend (José's work)

# Test the backend
uv run python test_backend.py

# Start the API server
uv run python run_backend.py

Visit http://localhost:8000/docs for API documentation.

3. Available Operations

The system currently supports:

1. Scale Demand: Modify electricity demand by a factor

   • Example: "Increase demand by 20%"
   • Example: "Increase Houston demand by 30% in summer 2022"

2. Add Assets: Add new power plants

   • Example: "Add a 500MW solar plant in West Texas"

Project Structure

• backend/: FastAPI app with Anthropic Claude integration
  • models.py: Pydantic data models
  • llm_interface.py: Claude integration for query interpretation
  • main.py: FastAPI application
• operations/: Pandas operations for CSV manipulation (Julian's work)
• frontend/: Streamlit chat interface (Toni's work)

Development

Code Quality

# Format code
uv run black .
uv run isort .

# Lint code
uv run ruff .

# Run pre-commit hooks
uv run pre-commit run --all-files