Utterly Voice Editor (Beta)

The Utterly Voice Editor is a voice optimized text and code editor with AI integration. You can use it for editing plain text, editing code, editing Utterly Voice settings files, or interacting with generative AI. Generative AI can significantly reduce the amount of typing you need to perform, so AI has been integrated into the editor. However, this AI integration is optional, so you can use the editor without AI.

Many popular AI user interfaces are not keyboard-friendly, so they are difficult to optimize by voice. In addition, most text/code editors are either too complicated for non-programmers, or they have too many actions that require using the mouse. The Utterly Voice Editor aims to fix these problems for voice-only users.

This editor is not an IDE. The user interface takes a minimalist approach, similar to a terminal-based editor with no menus or buttons. The AI integration is designed for users who prefer micromanaging AI, rather than letting AI have full project/computer control. All AI file reads/writes are only performed with your explicit review and approval using a voice driven change-by-change accept/reject process.

This page covers many of the basic commands, but you can see the details of all commands by saying "open help editor".

Beta Testing

This editor is currently in beta testing, so please report any issues you discover or feature requests via one of the channels on the help page.

The following are known issues, which will be fixed in the next release:

• The editor currently does not support display scaling other than 100%, which is commonly used for high pixel density displays.

Start the Editor

You can open the editor by saying "open utterly editor", which is a command in the "windows" mode. This will execute utterlyeditor.exe found in the Utterly Voice application folder. This command also activates the "editor" mode. You can open multiple editor windows using this command.

After it starts, you should say "maximize window", so the editor has plenty of screen space.

To close the editor, say "close window" while the editor is active.

Panels

Watch panels demo

The editor has fixed major panels on the left and right side of the window. These major panels are used for editing content and interacting with AI. There is also a minor panel at the bottom of the window, which is used for simple input and status. You can use the commands "left panel" and "right panel" to move focus between these panels. If you are in the middle of certain multi-step actions that require focus on a particular panel, these commands are ignored.

Traditional text editors use tabs when opening many files, but this is not the most efficient approach when using voice. With the Utterly Voice Editor, you can open many files and folders, but only two of them can be displayed at once. To list all of your open files and folders, say "list panels". To display one of these panels, move the cursor up (example: "go up three") or down (example: "go down five") to position it anywhere on the line that contains the desired panel name, then say "submit".

Open a File

Watch file opening demo

To open a file, say "open navigator". This opens a navigator panel, which lets you browse the drives and folders on your computer. To open a folder or file, move the cursor to the line of the desired item, then say "submit". Notice the following heading and item types:

• Favorites: These are your favorite folders, which you can configure in the settings.
• Utterly Voice application folder: This is displayed, so you can easily navigate to your settings files.
• Drives: Each drive connected to your computer will be listed.
• New Folder: Select this item to create a new folder. When prompted, provide the folder name, then say "submit".
• New File: Select this item to create a new file. When prompted, provide the file name, then say "submit".
• ../: This goes up a folder in the hierarchy.

If you prefer to paste or type the complete path for a file, you can simply say "open file", paste or type the full path, then say "submit".

Enter Versus Submit

Many editors have varying behavior when you press enter, shift+enter, or control+enter. Sometimes this behavior is context dependent. The Utterly Voice Editor simplifies this by always having the same meaning for enter and control+enter. The "enter" command presses the enter key, always inserts a newline character, and does not submit any input. The "submit" command presses control+enter, does not insert a newline character, and submits user input.

Moving the Cursor

Watch cursor movement demo

The Utterly Voice Editor displays line numbers and column numbers for every 10th column with alternating background shading to aid navigation. To jump to a particular line, say "line" followed by the line number you want. To jump to a particular column, say "column" followed by the column number you want. For example:

• "line one fifty"
• "column thirty five"
• "line ten column twenty"

You can also use the default commands from the "basic" and "windows" modes. For example:

• "go up ten"
• "go down three"
• "go right seven"
• "go left six"
• "go home"
• "control home"
• "go end"
• "control end"
• "go left word three"
• "go right word four"

Select Text

Watch text selection demo

Most Windows editors require holding the shift key in order to make a selection. This becomes inefficient when using your voice, because you have to release the shift before switching to another window, and you may lose your selection in that process. The Utterly Voice Editor uses specific commands for starting and stopping selection that do not involve holding the shift key. In order to avoid conflicting with default commands, the editor commands use the word "pick" instead of "select". To start selection, say "start picking". To finish selection, say "finish picking", "cut" or "copy".

You can select characters to the left or right of the cursor:

• "pick left three"
• "pick right five"

You can select words to the left or right of the cursor:

• "pick left word three"
• "pick right word five"

You can select one or more lines, starting from the line of the cursor:

• "pick line"
• "pick line five"

Modify and Copy Text

Watch modify and copy text demo

The following basic cut copy and paste commands are available:

• "cut"
• "copy"
• "paste"

The following undo and redo commands are available:

• "scratch"
• "redo"

Basic backspace and delete commands are available:

• "backspace"
• "delete"
• "backspace three"
• "delete seven"

Forward and backward word deletion commands are available:

• "clear left word"
• "clear right word"
• "clear left word six"
• "clear right word five"

In addition to the commands described in this section, most of the commands in the "advanced" mode also work for the editor.

Find and Replace

Watch find and replace demo

• "find"
• say the text
• "submit"
• "next" to proceed to the next match
• "previous" to go back to the previous match
• "cancel command" to stop finding
• "finish picking" to unselect the final match

• "replace"
• say the text you want to replace
• "submit"
• say the replacement text
• "submit"
• "confirm" to accept a replacement and proceed to the
• "skip" to skip over the current replacement
• "confirm all" to accept all replacements
• "cancel command" to stop replacing

When providing text for find or replace, you can include newline or tab characters by saying "enter" or "tab". The bottom panel that accepts this input only displays a single line, but it will display special characters for this input.

Save a File

Watch file saving demo

If you open a file or use the navigator to create a new file, then make changes to it, you can save the file by saying "save file".

If you want to save a file as a different name in the same folder (or in a different folder which you can type the path to), say "save as" then provide the new path and file name.

If you want to save a file in a different folder, and you do not want to type the path to the new file:

• "save navigate"
• navigate to the desired folder for the new file
• position the cursor at New File
• "submit"
• provide the file name
• "submit"

Configure Gemini (Optional)

If you configure Google's Gemini for the editor, you can use it to have an AI chat, read and explain files, or update files. You need to get a Gemini API key and configure the editor to use it. You can use the model for free, but if you use it often, you should sign up for a pay-as-you-go option.

Note that Google may save prompts and responses for review and model training, so you should understand Google's privacy agreement before sending confidential information to Gemini.

1. Follow these steps to get an API key from Google. State that you are an individual developer (not an organization). Set up billing for improved model performance.
2. Open the config\editor.yaml file in the Utterly Voice application directory.
3. Scroll down and find the gemini settings, then apiKeySecret within the Gemini settings. Update the string value to contain your Gemini API key. Keep this key private and secure; if somebody acquires your key, you will be billed for their usage.
4. Save the file.

You can manage your keys later by visiting the AI Studio.

Gemini Chats

Watch Gemini chats demo

To open a Gemini panel, say "open gemini". This opens a panel similar to a command line terminal. You can type input asking Gemini a question.

To have a chat with Gemini, just say your question, then say "submit". Gemini will respond, and you can provide the next input. For example, type "why is the sky blue?", followed by "submit".

Gemini maintains context through the conversation, so you can follow up with a question "how about the ocean?", and Gemini will explain why the ocean is blue.

Gemini File Reading and Writing

Watch Gemini file reading and writing demo

The Utterly Voice Editor does not provide Gemini direct access to any files on your computer. It can only read and write files that you have opened in the editor. When Gemini attempts to change a file, it does not actually change the file on disk. The editor will display the changes Gemini is suggesting, and it is up to you to accept or reject these changes and to save the file once completed.

To ask Gemini to read or modify a file, display the file in the active panel, then say "gemini read" or "gemini modify". You can repeat this process for multiple files. Each time you do this, the editor queues the read/modify requests. Once you are done flagging files for reading or modifying, say "open gemini". Input for Gemini to perform reads and modifications will be automatically inserted in the Gemini prompt. Follow this inserted content with your question or directions, then say "submit".

If Gemini has modified content, it typically provides an overview of what has changed. You can switch to each panel, then review/accept/reject the changes. For each change, the following markers are inserted in the panel content:

• <<< OLD >>>: marks the beginning of the old content that Gemini is replacing
• <<< NEW >>>: marks the beginning of the new content that Gemini is providing
• <<< END >>>: marks the end of this change

To review/accept/reject changes:

• "next" to jump to the next change
• "previous" to jump to the previous change
• "confirm" to accept a change and proceed to the next
• "confirm three" to accept the next three changes
• "confirm all" to accept all changes
• "skip" to reject a change and proceed to the next
• "skip five" to reject the next five changes
• "reload panel" to reject all changes and reload the file from disk
• "save file" to save the changes to disk

Gemini URL Context

Watch Gemini URL context demo

Gemini can read and answer questions about webpages. You just need to provide the URL in your question. For example: "Explain the content at utterlyvoice.com".

Gemini For Editing Utterly Voice Settings Files

Watch Gemini editing settings files demo

You can use Gemini for editing Utterly Voice settings files. The "editor" mode includes a special command called "gemini reference utterly", which inserts an instruction to read important Utterly Voice documentation files on customizing settings. To use Gemini for editing a settings file:

• Open and mark a settings file for modification as described above.
• Say "open gemini".
• Say "gemini reference utterly".
• Instruct Gemini on how to modify the file. For example, "add a file save command".
• Say "submit".

Customize the Editor

There are many customization settings for the editor. You can change font, font size, colors of various items, Gemini configuration, code formatting, etc. See Editor Settings.