pyfun-ai 0.1.0

LLM-powered local website builder with realistic generation workflow and delightful UX

Pyfun AI

Pyfun AI is an LLM-powered local website builder for developers who want rapid idea-to-UI generation directly from the terminal.

It simulates a realistic multi-step AI generation pipeline, creates a production-style landing page, and serves it instantly on localhost.

🚀 Quick Start

# Install globally
pip install git+https://github.com/yourusername/pyfun_ai.git

# Create a new project
pyfun init my_awesome_site
cd my_awesome_site

# Generate and preview
pyfun run --prompt "Build a modern SaaS landing page"

# Your browser opens automatically with your generated site!

That's it. Your site is live at http://localhost:8000 with AI-generated layout, features, and styling.

Why Pyfun AI

• Fast local workflow with zero external API dependency
• Believable AI-like generation logs and confidence scoring
• Clean, long-form HTML output with responsive sections
• Built-in project initialization and one-command run experience
• Use as CLI tool or import as Python library

Installation

Option 1: Install from GitHub (Global)

pip install git+https://github.com/yourusername/pyfun_ai.git

Option 2: Install from PyPI (when published)

pip install pyfun_ai

Option 3: Install for Development

Clone the repository and install in editable mode:

git clone https://github.com/yourusername/pyfun_ai.git
cd pyfun_ai
pip install -e .

This allows you to modify code and test changes immediately without reinstalling.

Usage

1) Initialize a project

pyfun init my_site
cd my_site

2) Generate and run

pyfun run --prompt "Create a modern SaaS launch page" --port 8000

Then open the local URL shown in terminal.

CLI Commands

• pyfun init <project_name>

  • Creates a new project directory
  • Adds pyfun.json metadata
  • Prepares output directory scaffold

• pyfun run [--prompt "..."] [--port 8000]

  • Simulates LLM generation and scoring
  • Builds site/index.html
  • Starts a local server and opens browser

Use as a Library

You can integrate pyfun_ai into your own Python projects:

Example 1: Generate Website Programmatically

from pathlib import Path
from pyfun_ai.core.composer import generate_website_config
from pyfun_ai.core.builder import generate_website

# Generate layout configuration
result = generate_website_config("Create a modern SaaS landing page")

# Get the template
package_root = Path(__file__).resolve().parent
template_path = package_root / "pyfun_ai" / "templates" / "base.html"

# Generate HTML file
output_dir = Path("./my_site/site")
index_file = generate_website(result.config, output_dir, template_path)

print(f"✅ Site generated at: {index_file}")

Example 2: Custom Generation with Theme Control

from pyfun_ai.core.composer import THEME_PALETTES

# Access available themes
print("Available themes:", list(THEME_PALETTES.keys()))
# Output: Available themes: ['modern', 'minimal', 'neon']

# Generate and use theme colors
result = generate_website_config("Build a portfolio site")
theme_config = result.config["theme_palette"]
print(f"Primary accent color: {theme_config['accent']}")

Example 3: Access Core Modules

from pyfun_ai.core.composer import (
    generate_website_config, 
    COMPONENT_FEATURES, 
    OPTIMIZATION_HINTS
)
from pyfun_ai.utils.terminal import print_progress, print_staged

# Use generation pipeline
result = generate_website_config("Your prompt here")
print(f"Quality Score: {result.confidence_score}%")

# Access feature bank
print(f"Available components: {len(COMPONENT_FEATURES)}")

# Use terminal utilities
print_progress("Building your site", 1.5)
print_staged("Your site is ready!", 0.8)

Module API Reference

pyfun_ai.core.composer

• generate_website_config(prompt: str) -> CompositionResult
  • Returns a config dict and confidence score
  • Prints realistic AI optimization logs to stdout

pyfun_ai.core.builder

• generate_website(config, output_dir, template_path) -> Path
  • Renders HTML from template with injected config
  • Returns path to generated index.html

pyfun_ai.core.server

• run_project(project_dir, prompt, port) -> None
  • Full pipeline: composer → build → serve → open browser

pyfun_ai.utils.terminal

• print_progress(label, delay_seconds)
  • Print animated progress bar
• print_staged(text, delay_seconds)
  • Print message with delay for effect

Feature Overview

• Prompt-compatible random generation
  • Generates a fresh randomized UI every run, while simulating prompt understanding
• Structured long-form website sections
  • Hero, feature grid, metric cards, testimonials, and polished footer
• Generation effects
  • Progress-bar style logs, confidence score, optimization mode, and AI hints
• Theme engine simulation
  • Rotates modern, minimal, and neon style palettes per generation
• Local serving
  • Uses Python standard library server for portability and instant preview

Generated Output

Pyfun AI writes generated assets to:

• ./site/index.html

Screenshots (mock descriptions)

1. Terminal Generation Pipeline
   • Startup engine checks, AI progress stages, confidence score, and optimization mode
2. Generated Website Hero
   • Bold heading, trust badge, mode tag, and preset label
3. Feature + Stats Section
   • Card-based feature grid plus real-time looking generation metrics
4. Social Proof Block
   • Testimonial cards showing polished product storytelling

Security & Privacy

• No telemetry
• No tracking
• No user-data collection
• No external AI API calls

Requirements

• Python 3.9+

Publishing to PyPI

To publish pyfun_ai to PyPI and make it globally installable:

# Install build tools
pip install build twine

# Build distribution
python -m build

# Upload to PyPI Test (recommended first)
twine upload --repository testpypi dist/*

# Upload to PyPI Production
twine upload dist/*

Then users can install globally with:

pip install pyfun_ai

Contributing

Pull requests welcome. For major changes, please open an issue first.

License

MIT