Command Execution
myHand.ai goes beyond just being a chatbot by offering a unique and powerful capability - the ability to execute commands and perform computing tasks on your behalf. Unlike a mere chatbot, myHand.ai can interact with your computer system and carry out specific commands to accomplish various computing tasks. This feature allows you to leverage the expertise and efficiency of myHand.ai to automate processes, streamline workflows, and perform complex tasks with ease. However, it is essential to remember that with great power comes great responsibility, and users should exercise caution and use this feature at their own risk.
Command execution helps:
- to get the requested information, e.g.
tell me the current time
- to execute computing tasks on local device, e.g.
go to my Desktop
list all files with names started with "Screenshot"
delete them
- to interact with third-party applications, e.g.
open "my_music.mp3" with VLC player
open Safari and search for "ChatGPT"
- to interact with os assistant, e.g.
open Siri
Please be aware that myHand.ai is a virtual assistant that has the ability to execute commands and perform computing tasks on your behalf. While this feature can be powerful and convenient, it is important to understand that any changes made to your computer's system or data are done at your own risk.
The developer of myHand.ai cannot be held liable for any negative impact or consequences that may occur as a result of using the command execution feature inappropriately or without proper knowledge and understanding. It is recommended that you exercise caution and ensure you fully understand the implications of the commands you request myHand.ai to execute.
By using the command execution feature of myHand.ai, you acknowledge and accept full responsibility for any potential risks or consequences that may arise from the execution of these commands.
myHand.ai has two modes of command execution, "enhanced" and "auto". The major difference between the two modes is that the "enhanced" mode always prioritizes command execution over a chat response.
In the "enhanced" mode, command execution is given the highest priority. Whenever a user enters a request, myHand.ai screens for available commands and executes them if possible. This ensures that commands are executed promptly.
In the "auto" mode, ChatGPT automatically decides whether command execution may be helpful. This means that myHand.ai uses its internal intelligence to determine if executing a command can provide assistance or additional information to the user.
These two levels of command execution allow myHand.ai to effectively respond to user requests, both through chat-based interactions and by executing relevant commands when appropriate.
To make a change:
-
enter a "blank" entry, by pressing the "Enter" key
-
select "change command execution mode
-
select either "enhanced" or "auto"
-
click "ok"
You can use keyboard shortcuts, "ctrl + e", to quickly swap between the "enhanced" and the "auto modes.
myHand.ai offers a command execution feature that allows users to execute system commands and Python code through the assistant. However, executing commands can carry certain risks, such as file deletions or other significant impacts. To provide users with control over these risks, myHand.ai offers four different options to manage the risk associated with command execution. These options determine when a confirmation prompt is displayed before a command is executed.
Before diving into the options, it's important to understand the risk assessment levels used by myHand.ai. The risk assessment levels are categorized as follows:
-
High Risk: This level includes commands that have the potential to cause significant impacts, such as file deletions or other irreversible actions.
-
Medium Risk: This level includes commands that may have some potential risks or impacts, but they are not as severe as high-risk commands.
-
Low Risk: This level includes commands that have minimal or no potential risks or impacts.
-
Always: With this option, a confirmation prompt is displayed for every command execution, regardless of the risk level. Users are required to confirm the execution of the command every time.
-
Medium Risk or Above: This option prompts a confirmation prompt only when the risk level of the command is medium or high. If the risk level is low, the command is executed without any prompt.
-
High Risk Only: With this option, a confirmation prompt is displayed only when the risk level of the command is high. If the risk level is medium or low, the command is executed without any prompt.
-
None: This option disables the confirmation prompt entirely. Users can execute commands without any prompts, regardless of the risk level. It is important to note that choosing this option means users take full responsibility for the potential risks associated with command execution.
The choice of the risk management option depends on the user's preference for control and their comfort level with potential risks. Here are some considerations for choosing the right option:
-
High Level of Control: If users want maximum control over command execution and want to review and confirm every command, they should choose the "Always" option.
-
Balanced Control: If users want to have control over medium and high-risk commands but are comfortable with low-risk commands executing without prompts, they should choose the "Medium Risk or Above" option.
-
Minimal Control: If users want to minimize prompts and are only concerned about high-risk commands, they should choose the "High Risk Only" option.
-
Full Trust: If users fully trust the commands they execute and want a seamless experience without any prompts, they can choose the "None" option. However, it is important to exercise caution and ensure the commands being executed are safe.
It is recommended to carefully consider the risks associated with command execution and choose the option that aligns with the user's preferences and level of comfort.
To make a change:
-
enter a "blank" entry, by pressing the "Enter" key
-
select "change user confirmation
-
select one of the options
-
click "ok"
You may also choose to display the command before they are executed.
To view the code before execution, enable either 'command display' or 'developer mode'.
