AI Tools GUI Automation

IMPORTANT DISCLAIMER > Use with your own caution. You can be banned because of an automation it's unclear. This tool automates web-based AI platforms (ChatGPT, Gemini, Claude) which may violate their Terms of Service. Using automation could result in account suspension or permanent bans. The authors are not responsible for any consequences. Use at your own risk.

Automate bulk artifact generation (images, text, code) across multiple AI platforms using Selenium-based browser automation.

What This Tool Does

This automation tool eliminates the tedium of manually generating hundreds of artifacts using AI web interfaces.

Without This Tool:

Open ChatGPT/Gemini/Claude manually
Type each prompt one by one
Wait for generation
Download each file individually
Rename and organize files manually

With This Tool:

Define all artifacts in a markdown file
Run one command: python src/main.py prompts.md
Let automation handle everything
Get organized artifacts in ./artifacts folder

Features

Multi-Provider Support: Works with ChatGPT, Gemini, and Claude
Bulk Generation: Process hundreds of artifacts automatically
Smart Download Management: Automatic file detection, waiting, and organization
Chrome Profile Support: Stay logged in or use your existing profile
Flexible Configuration: JSON-based configuration system
Error Handling: Automatic retries, detailed logging, error screenshots
Progress Tracking: Real-time progress indicators and statistics
Session Management: Reuses browser sessions across artifacts
Type Support: Images, text documents, code files
Skip Existing: Optionally skip already-generated artifacts
Headless Mode: Run invisibly in the background

Project Structure

ai-tools-gui-automation/
├── src/                            # Source code
│   ├── __init__.py                # Package initialization
│   ├── main.py                    # Main orchestration script
│   ├── markdown_parser.py         # Parse artifact definitions
│   ├── file_manager.py            # File operations and organization
│   ├── base_provider.py           # Base class for AI providers
│   ├── gemini_provider.py         # Gemini automation
│   ├── chatgpt_provider.py        # ChatGPT automation
│   └── claude_provider.py         # Claude automation
│
├── bulk_data/                      # Your markdown prompt files
│   └── EXAMPLE_PROMPTS.md          # Example: 76 prompts
│
├── artifacts/                      # Generated artifacts (gitignored)
├── downloads/                      # Temporary downloads (gitignored)
├── chrome_automation_profile/      # Dedicated Chrome profile (gitignored)
│
├── config.json                     # Configuration settings
├── requirements.txt                # Python dependencies
├── test_setup.py                   # Setup verification script
├── find_chrome_profile.py          # Find your Chrome profile path
│
├── PROJECT_SUMMARY.md              # Detailed project documentation
├── QUICK_START.md                  # Quick start guide
├── example_usage.md                # Usage examples
└── README.md                       # This file

Installation

Prerequisites

Python 3.8+
Google Chrome (latest version)
Active accounts on AI platforms you want to use (ChatGPT, Gemini, Claude)

Step 1: Clone Repository

git clone https://github.com/yourusername/ai-tools-gui-automation.git
cd ai-tools-gui-automation

Step 2: Install Dependencies

pip install -r requirements.txt

Step 3: Verify Setup

python test_setup.py

This checks:

Python version
Dependencies installation
Chrome and ChromeDriver
Directory structure
Core functionality

Configuration

Edit config.json to customize behavior:

{
  "download_dir": "./downloads",
  "artifacts_dir": "./artifacts",
  "headless": false,
  "timeout": 300,
  "retry_attempts": 3,
  "delay_between_artifacts": 5,

  "chrome_profile": {
    "enabled": true,
    "use_existing_profile": false,
    "user_data_dir": ".\\chrome_automation_profile",
    "profile_directory": "Default"
  },

  "providers": {
    "gemini": {
      "enabled": true,
      "url": "https://gemini.google.com/app"
    },
    "chatgpt": {
      "enabled": true,
      "url": "https://chat.openai.com"
    },
    "claude": {
      "enabled": true,
      "url": "https://claude.ai"
    }
  },

  "file_handling": {
    "create_backups": true,
    "skip_existing": true,
    "clear_downloads_after": true
  },

  "logging": {
    "level": "WARNING",
    "file": "automation.log"
  }
}

Key Configuration Options:

Option	Description	Default
`download_dir`	Where browsers download files	`./downloads`
`artifacts_dir`	Where to save organized artifacts	`./artifacts`
`headless`	Run browsers invisibly	`false`
`timeout`	Max wait time for generation (seconds)	`300`
`retry_attempts`	Retries on failure	`3`
`delay_between_artifacts`	Delay between generations (seconds)	`5`
`skip_existing`	Skip already-generated artifacts	`true`

Usage

Basic Usage

python src/main.py bulk_data/EXAMPLE_PROMPTS.md

With Filters

# Only process Gemini artifacts
python src/main.py prompts.md --filter-provider gemini

# Only process images
python src/main.py prompts.md --filter-type image

# Only process text documents
python src/main.py prompts.md --filter-type text

Advanced Options

# Headless mode (no browser window)
python src/main.py prompts.md --headless

# Regenerate everything (don't skip existing)
python src/main.py prompts.md --no-skip-existing

# Custom config file
python src/main.py prompts.md -c custom_config.json

# Combine options
python src/main.py prompts.md --filter-provider gemini --headless

Command-Line Arguments

usage: main.py [-h] [-c CONFIG] [--headless] [--no-skip-existing]
               [--filter-provider {gemini,chatgpt,claude}]
               [--filter-type {image,text,code}]
               markdown_file

positional arguments:
  markdown_file         Path to markdown file with artifact definitions

optional arguments:
  -h, --help            Show help message
  -c CONFIG, --config CONFIG
                        Path to config file (default: config.json)
  --headless            Run in headless mode
  --no-skip-existing    Regenerate all artifacts, even if they exist
  --filter-provider     Only process artifacts for specific provider
  --filter-type         Only process artifacts of specific type

Markdown Format

Define your artifacts in a markdown file using one of two formats:

Structured Format (Recommended)

### Luck Potion Icon

**Type:** image
**Provider:** gemini
**Output Name:** luck_potion_icon
**Extension:** png

Create a vibrant RPG game icon for a luck potion.
The icon should be 512x512 pixels with a transparent background.
Show a glass bottle filled with glowing green liquid and a four-leaf clover symbol.
Art style: Cartoon/stylized, bright colors, clear outlines.

Simple Format (Quick & Easy)

### Luck Potion Icon

Create a vibrant RPG game icon for a luck potion.

Auto-assumes: type=image, provider=gemini, extension=png

Metadata Fields

Field	Required	Description	Default
`Type`	No	`image`, `text`, or `code`	`image`
`Provider`	No	`gemini`, `chatgpt`, or `claude`	`gemini`
`Output Name`	No	Filename without extension	Generated from artifact name
`Extension`	No	File extension	`png` for images, `txt` for text/code

Example File Structure

See bulk_data/EXAMPLE_PROMPTS.md for a complete example with 76 game asset definitions.

Supported AI Providers

Gemini (Google)

Best for: Image generation (Imagen 3)
Supported types: Images, text
URL: https://gemini.google.com/app
Notes: Fast image generation, high quality

ChatGPT (OpenAI)

Best for: DALL-E images, text generation
Supported types: Images (DALL-E), text
URL: https://chat.openai.com
Notes: Requires ChatGPT Plus for DALL-E

Claude (Anthropic)

Best for: Text, code generation
Supported types: Text, code
URL: https://claude.ai
Notes: No native image generation

How It Works

Step-by-Step Process

Parse Markdown: Extract artifact definitions from markdown file
Initialize Providers: Set up browser automation for needed providers
For Each Artifact:
- Get or create provider instance
- Clear download directory
- Select generation mode (image/text/code)
- Send prompt to AI platform
- Wait for completion
- Download artifact
- Rename and organize file
Handle Errors: Retry on failure, save error screenshots
Generate Report: Summary of successful and failed artifacts

Smart Download Detection

The tool monitors your download directory and:

Waits for new files to appear
Checks file size is stable (not still downloading)
Filters out temporary files (.crdownload, .tmp)
Renames with meaningful names
Moves to artifacts folder

Session Management

Keeps browsers open between artifacts
Reuses login sessions
Only one browser instance per provider
Logs in once, generates many artifacts

Chrome Profile Options

Choose one of two approaches:

Option A: Use Your Existing Chrome Profile (No Login Required)

Pros: Already logged in to all AI platforms Cons: Must close Chrome while automation runs

{
  "chrome_profile": {
    "enabled": true,
    "use_existing_profile": true
  }
}

Steps:

Close all Chrome windows
Run automation
Already logged in!

Find your Chrome profile path:

python find_chrome_profile.py

Option B: Dedicated Automation Profile (Recommended)

Pros: Runs alongside your regular Chrome, stays logged in Cons: Must log in once on first run

{
  "chrome_profile": {
    "enabled": true,
    "use_existing_profile": false,
    "user_data_dir": ".\\chrome_automation_profile"
  }
}

First time:

Run automation
Chrome opens, log into AI platforms
Automation continues

Every time after: Just run - stays logged in!

🐛 Common Issues

Issue: ChromeDriver Not Found

Solution: Automatic via webdriver-manager. If fails, manually download ChromeDriver matching your Chrome version.

Issue: Login Fails

Solution:

Increase timeout in config
Manually log in when browser opens
Use Chrome profile option to stay logged in

Issue: Elements Not Found

Solution: AI platform UI changed. Update selectors in provider code (src/*_provider.py).

Issue: Downloads Not Detected

Solution:

Check download directory exists and has write permissions
Ensure Chrome downloads to correct folder
Check config.json download_dir path

Issue: Rate Limiting

Solution:

Increase delay_between_artifacts in config
Use --filter-type or --filter-provider to process fewer artifacts
Spread generation across multiple days

Issue: Account Banned

Remember: This tool violates most AI platforms' Terms of Service. Use at your own risk.

Performance & Optimization

Speed Optimization

Headless Mode: 20-30% faster

python src/main.py prompts.md --headless

Skip Existing: Only generate new artifacts

"file_handling": {
  "skip_existing": true
}

Reduce Delay: Decrease wait time between artifacts (risk rate limiting)
```
"delay_between_artifacts": 2
```

Resource Usage

Memory: ~200-500MB per browser instance
Disk: Downloads cleaned up automatically
Network: Varies by artifact size

Scalability

Can process hundreds of artifacts
Limited by AI platform rate limits
Can distribute across multiple machines with separate accounts

Best Practices

Start Small: Test with 2-3 artifacts first
Monitor First Run: Don't use headless mode initially
Clear Prompts: Detailed prompts = better results
Use Chrome Profiles: Avoid repeated logins
Check Logs: Review automation.log for issues
Respect Rate Limits: Don't overwhelm AI platforms
Backup Important Work: Keep copies of critical artifacts
Version Control: Track markdown files, not artifacts (add to .gitignore)
Test Providers: Verify each provider works before bulk generation
Read ToS: Understand risks of automation

Extending the Tool

Add a New AI Provider

Create src/new_provider.py:

from base_provider import BaseAIProvider

class NewProvider(BaseAIProvider):
    def __init__(self, download_dir, headless=False):
        super().__init__(download_dir, headless)
        self.base_url = "https://newai.example.com"

    def login(self, credentials=None):
        # Implement login logic
        self.driver.get(self.base_url)
        # Handle authentication

    def select_mode(self, mode):
        # Implement mode selection (image/text/code)
        pass

    def send_prompt(self, prompt):
        # Implement prompt submission
        pass

    def wait_for_completion(self, timeout=300):
        # Implement waiting for AI to finish
        pass

    def download_artifact(self, artifact_name):
        # Implement download logic
        pass

Register in src/main.py:

from new_provider import NewProvider

# In get_provider method:
elif provider_name == "newprovider":
    provider = NewProvider(
        download_dir=self.config.get("download_dir"),
        headless=self.config.get("headless", False)
    )

Update config.json:

"providers": {
  "newprovider": {
    "enabled": true,
    "url": "https://newai.example.com"
  }
}

Customize Processing Logic

Extend the orchestrator:

class CustomOrchestrator(AIAutomationOrchestrator):
    def process_artifact(self, artifact):
        # Custom pre-processing
        print(f"Processing with custom logic: {artifact.name}")

        result = super().process_artifact(artifact)

        # Custom post-processing
        if result:
            self.apply_custom_filters(artifact)

        return result

Dependencies

selenium (4.16.0): Web automation framework
webdriver-manager (4.0.1): Automatic ChromeDriver management
pathlib2 (2.3.7): Enhanced path handling
python-dotenv (1.0.0): Environment configuration
tqdm (4.66.1): Progress bars
colorama (0.4.6): Colored terminal output

License

MIT License

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

Final Warning

Use with your own caution. You can be banned because of an automation it's unclear.

This tool automates interactions with web-based AI platforms, which may violate their Terms of Service. Potential consequences include:

Account suspension or permanent ban
Loss of paid subscriptions
IP address blocking
Legal action (unlikely but possible)

The authors of this software:

Do NOT encourage violating Terms of Service
Are NOT responsible for any consequences
Provide this tool for educational purposes only
Recommend reviewing each platform's ToS before use

Use at your own risk. You have been warned.

Additional Documentation

PROJECT_SUMMARY.md: Detailed technical documentation
QUICK_START.md: Quick start guide for Chrome profiles
example_usage.md: Usage examples and patterns

Contributing

Contributions welcome! Please:

Fork the repository
Create a feature branch
Commit your changes
Push to the branch
Open a Pull Request

Support

Issues: Check automation.log for detailed error information
Screenshots: Error screenshots saved to downloads directory
Documentation: See additional markdown files in repository
Testing: Run python test_setup.py to verify setup

Built with Python and Selenium for automated AI artifact generation

Remember: Use responsibly and at your own risk!

Name		Name	Last commit message	Last commit date
Latest commit History 9 Commits
src		src
.gitattributes		.gitattributes
.gitignore		.gitignore
README.md		README.md
SETUP.md		SETUP.md
TEMPLATE.md		TEMPLATE.md
config.json		config.json
find_chrome_profile.py		find_chrome_profile.py
quickstart.bat		quickstart.bat
quickstart.sh		quickstart.sh
requirements.txt		requirements.txt
test_setup.py		test_setup.py

tugcantopaloglu/ai-tools-gui-automation

Folders and files

Latest commit

History

Repository files navigation

AI Tools GUI Automation

Table of Contents

What This Tool Does

Without This Tool:

With This Tool:

Features

Project Structure

Installation

Prerequisites

Step 1: Clone Repository

Step 2: Install Dependencies

Step 3: Verify Setup

Configuration

Key Configuration Options:

Usage

Basic Usage

With Filters

Advanced Options

Command-Line Arguments

Markdown Format

Structured Format (Recommended)

Simple Format (Quick & Easy)

Metadata Fields

Example File Structure

Supported AI Providers

Gemini (Google)

ChatGPT (OpenAI)

Claude (Anthropic)

How It Works

Step-by-Step Process

Smart Download Detection

Session Management

Chrome Profile Options

Option A: Use Your Existing Chrome Profile (No Login Required)

Option B: Dedicated Automation Profile (Recommended)

🐛 Common Issues

Issue: ChromeDriver Not Found

Issue: Login Fails

Issue: Elements Not Found

Issue: Downloads Not Detected

Issue: Rate Limiting

Issue: Account Banned

Performance & Optimization

Speed Optimization

Resource Usage

Scalability

Best Practices

Extending the Tool

Add a New AI Provider

Customize Processing Logic

Dependencies

License

Final Warning

Additional Documentation

Contributing

Support

About

Topics

Resources

Uh oh!

Stars

Watchers

Forks

Contributors

Uh oh!

Languages