Rails Superhero Cards

Get an AI-generated Rails superhero card from a set of skills and a photo.

sequenceDiagram
    User->>Frontend: Click Generate
    Frontend->>Backend: POST /generate-hero-card
    Backend->>Celery: Queue task
    Backend-->>Frontend: 202 + task_id
    Frontend->>Frontend: Start polling
    loop Every 5 seconds
        Frontend->>Backend: GET /status/{task_id}
        Backend->>Celery: Check task status
        Celery-->>Backend: Status + result
        Backend-->>Frontend: Status response
    end
    Frontend-->>User: Display card when complete

Loading

Local Setup

Key dependencies:

• Python 3.12
• FastAPI 0.124.2
• Node 22
• React 19
• Postgres 16
• Localstack

With Taskfile

Download the task runner:

brew install go-task/tap/go-task

Copy .env.example and fill in your environment variables.

Run task up:build to build the Docker images and start the application.

Run task migrate to run migrations.

Navigate to http://localhost:3003 to access the application.

To run the linter do task lint. Or for specific linters:

task lint:backend
task lint:frontend

Run task to see a list of all available tasks.

With Docker Compose

Copy .env.example and fill in your environment variables.

Run docker-compose up -d to start the application.

Run docker-compose exec backend uv run alembic upgrade head to run migrations.

Navigate to http://localhost:3003 to access the application.

Langfuse Setup

The app uses Langfuse to trace LLM conversations. You can set up an account at https://cloud.langfuse.com.

By default, tracing is off. Set the ENABLE_LANGFUSE environment variable to true to enable it.