README.md

metadata

title: Gotti Signal Gen
emoji: 🏢
colorFrom: green
colorTo: gray
sdk: docker
pinned: false

Stock Alchemist Signal Generator

A Dockerized FastAPI application that generates stock trading signals using Ollama (LLaMA 3.1) and performs scheduled fundamental analysis. Designed to run on Hugging Face Spaces.

Features

• Daily Signal Generation: Prompts a local LLaMA 3.1 model to analyze stocks and generates trading signals.
• Weekly Analysis: Runs a comprehensive fundamental analysis script every Saturday.
• Persistent Storage: Stores results in a TiDB/MySQL database.
• Automated Scheduling: Uses GitHub Actions to trigger daily and weekly tasks.

Setup Instructions

1. Hugging Face Space Setup

1. Create a new Space on Hugging Face.
2. Select Docker as the SDK.
3. Choose "Blank" or just point it to your repository.

2. Environment Variables

In your Hugging Face Space "Settings" -> "Variables and secrets", add the following Secret variables:

• API_SECRET: A strong secret key to protect your endpoints.
• DB_HOST: Your TiDB/MySQL host.
• DB_PORT: Your TiDB/MySQL port (e.g., 4000 for TiDB, 3306 for MySQL).
• DB_USERNAME: Database username.
• DB_PASSWORD: Database password.
• DB_DATABASE: Database name (e.g., gotti).
• DB_SSL_CA: (Optional) Path to SSL CA certificate if needed (relative to repository root).

3. GitHub Actions Setup

To enable the scheduler:

1. Go to your GitHub repository -> Settings -> Secrets and variables -> Actions.
2. Add the following Repository secrets:
   • HF_SPACE_URL: The URL of your Hugging Face Space (e.g., https://username-space-name.hf.space).
   • API_SECRET: The same secret you set in Hugging Face.

4. Database Initialization

The application handles table creation automatically in src/db/local_database.py. Ensure your database user has CREATE TABLE permissions.

Local Development

Prerequisites

• Docker
• Python 3.10+
• Ollama (running locally if testing outside Docker)

Running with Docker

# Build the image
docker build -t stock-alchemist .

# Run the container
# Note: You need to pass env vars. For local testing, you might want to use --env-file
docker run -p 7860:7860 \
  -e API_SECRET=test_secret \
  -e DB_HOST=host.docker.internal \
  -e DB_USERNAME=root \
  -e DB_PASSWORD=yourpassword \
  stock-alchemist

API Endpoints

• POST /generate-signal: Trigger signal generation. Headers: X-API-Secret: <your_secret> (or API_SECRET in local dev). Note: The code uses API_SECRET env var validation.
• POST /saturday-analysis: Trigger weekly analysis.
• GET /health: Health check and system vitals.

Project Structure

• Dockerfile & entrypoint.sh: System configuration.
• src/main.py: FastAPI application entry point.
• src/db/: Database interaction logic.
• .github/workflows/scheduler.yml: Automated scheduler.