VecchioGPT

System-wide GPT prompts computed on your clipboard's contents! Ghe provo però vecchio bel casino vero? Sisi, si

VecchioGPT is a Python application that allows users to apply custom GPT prompts to the text stored in the system clipboard. It is cross-platform and compatible with Windows, Linux and MacOS systems.

Demo video

OpenAI API

VecchioGPT runs on the OpenAI API, which means that you need to have an OpenAI account and an OpenAI API key. Unlike the web version of ChatGPT Plus, which operates on a monthly subscription basis, API access is billed based on usage at very affordable prices, giving you access to all OpenAI models and tools.

The workflow is pretty easy: you can copy some text to your system clipboard, call a custom prompt using your clipboard contents as context, then the result will be copied to the clipboard, ready to be pasted anywhere.

Features

Main features

• Clipboard Integration: Easily generate prompts which include the contents of your system clipboard.
• Customization: Easily create custom prompts to cover a wide range of use cases, using an extremely easy syntax.
• Sharing: Prompts are saved in JSON files which can be easily shared.
• Workflow Efficiency: Use a designated hotkey to call VecchioGPT, and the user-defined prompt is executed, with the output conveniently pasted into the clipboard.
• Interactive GUI: A cross-platform graphical user interface provides a user-friendly way to select and customize prompts.

Extras

• Voice command: Call a voice prompt with/without using your clipboard contents as context (using OpenAI Whisper)
• Read aloud: Speak the contents of your clipboard (using OpenAI TTS)

Use cases

• Easily modify your code using GPT. You can define your own workflow through system messages. For example, you can instruct a prompt to generate code from your comments
• Support your text writing process. Define some prompts which shorten text without losing informations, or turn a bad draft into well-written text.
• Summarize a long text.

Installation

0. Install Python on your machine. We developed VecchioGPT with Python 3.12, and we can't assure it will work on earlier versions.

1. Clone the repository (if you haven't done so already):

   git clone https://github.com/Neeqstock/VecchioGPT
   cd VecchioGPT

2. Create a virtual environment (optional but recommended):

   python -m venv venv

   Activate the virtual environment:

   • On Windows:

     venv\Scripts\activate

   • On macOS and Linux:

     source venv/bin/activate

3. Install the required packages:

   You can install all the necessary packages by running:

   pip install -r requirements.txt

4. (Optional, for Windows users) Add a start menu shortcut

   We included an icon, which you can find under the name VecchioGPT.ico. You can use that to make a custom link to VecchioGPT.bat and add it to your start menu links.

First run

In order to create the settings files, please run:

• CreateSettings.bat if you are on Windows
• CreateSettings.sh if you are on Linux

Then, create an OpenAI API key from your OpenAI API dashboard. Copy the OpenAI API key into the Settings.json file:

OPENAI_API_KEY=<your-OpenAI-key>

Usage

• Copy some text using CTRL+C
• Open the VecchioGPT prompt selection GUI
  • Navigate using TAB/SHIFT+TAB and arrow keys
  • Select the prompt from the list and edit the additional parameters, if necessary
• Launch the prompt by pressing ENTER
• You will hear two sounds: the first to indicate the computation has started, the last to indicate the computation has ended
• Paste the result using CTRL+V
• Prompts, input and answers will also be printed on the console

There are two ways to call the prompt selection GUI:

Method 1: background daemon with hotkeys

• Run VecchioGPT.bat if you are on Windows, or VecchioGPT.sh if you are on Linux. Read the indications in the console, which you can then reduce to icon if you'd like
• Press the default keyboard shortcut assigned to call the prompt selection GUI

Method 2: single launch

• Run VecchioGPT_GUI.bat if you are on Windows, or VecchioGPT.sh if you are on Linux

Sample prompts

When you will clone VecchioGPT, the prompts folder will be almost empty. In the Awesome-VecchioGPT folder you will find a list of curated prompts which you can copy in the prompts folder. You can take inspiration from these to create new ones!

Models

Model names can be edited in the gpt_models.json file. Sometimes OpenAI adds new ones, you can add them in this list.

Complexity

OpenAI released lots of different models with different context length, cost, and reasoning capacity. Since different prompts require different reasoning capacities, we implemented a way to define the required computational capacity for each prompt. Each prompt JSON file has a complexity field which can be set on three different values:

• low is for low-complexity computations, which require cheaper models with less reasoning power. If the complexity field is absent in the prompt JSON, this will be the default.
• high is for high-complexity computations, which require more expensive models with more reasoning power.
• long is for computation which require a high number of input and/or output tokens.

In order to explain VecchioGPT that is should take into account the complexity parameter, you should set file_specific as a model in the model settings. Otherwise, with any other specified model, the complexity parameter will be ignored, and that specific model will be used.

Settings

In the settings.json file you will find some tweakable settings. You can also edit the settings through the settings editor GUI (which you can call through shortcuts, or by running directly settingsGui.py).

• model: indicates the currently selected model
• OPENAI_API_KEY: this needs to be set in order to run VecchioGPT
• sound: enables/disables sound notifications (e.g. on prompt completion). Can be set to true or false
• popup: enables/disables popup notifications. Can be set to
  • off, i.e. no popup notifications
  • system, i.e. popups handled by the OS
  • vecchio, i.e. popups handled by VecchioGPT's visual popup system
• complexity-low_model: defines which model will handle the low complexity prompts
• complexity-high_model: defines which model will handle the high complexity prompts
• complexity-long_model: defines which model will handle the long complexity prompts
• speech-to-text_model: defines which model will handle speech-to-text calls
• text-to-speech_model: defines which model will handle text-to-speech calls
• text-to-speech_voice: defines which text-to-speech voice will be used

On OpenAI websites you can find a list of models and TTS voices.

Making new custom prompts

Creating new VecchioGPT prompts is very easy, through a very simple syntax. See the prompt creation guide. It's greatly advised not to create two prompts with the same promptName. In the templates folder you will find a sample prompt file.

Contribute

Feel free to contribute by forking the repository and creating pull requests. Bug reports and feature requests are welcome! However, we can't assure we will fulfill them :') VecchioGPT was created as a weekend project by two lazy dudes...

Funny extras

• We programmed the core functionalities of VecchioGPT in an evening using ChatGPT. When these were ready, we programmed the rest using VecchioGPT itself.
• Vecchio is a typical way to call a pal or a buddy in Veneto, an italian region.