prompt-optimizer-lite is a VS Code extension that helps you improve prompts without leaving your editor. It works with the chat model you already use in VS Code. If that fails, it falls back to a local Python optimizer, then to a built-in template.
Use it to:
- clean up rough prompts
- make instructions clearer
- shorten long text
- turn messy ideas into direct requests
- keep working even when the chat model does not respond
Visit this page to download and install the app package for Windows:
On the Releases page, look for the latest version and download the Windows file.
- Open the GitHub Releases page
- Download the latest Windows release file
- If the file is a ZIP archive, right-click it and choose Extract All
- Open the extracted folder
- If you see an installer, run it
- If you see a VS Code extension file, open VS Code and install it from the Extensions view
- Restart VS Code if asked
If Windows asks for permission, choose Yes.
Type @promptopt in VS Code Chat to optimize a prompt in conversation. This is the easiest way to use it when you already have Chat open.
Select text in the editor and run the optimize command. You can also right-click and use the context menu option.
If no text is selected, the extension opens an input box. Paste or type a prompt there and optimize it.
After optimization, you can:
- replace the selected text
- insert the result at the cursor
- copy the result to the clipboard
- open the result as a Markdown preview
prompt-optimizer-lite tries three paths in this order:
- current chat model
- local Python optimizer
- built-in template
This helps the tool keep working when one step fails.
- Open VS Code
- Open Chat
- Type
@promptopt - Enter your prompt
- Review the optimized result
- Open a file
- Select a prompt or note
- Run the optimize command
- Choose how you want the result applied
- Run the optimize command
- Type or paste your prompt in the box
- Pick where the result should go
You need:
- Windows 10 or Windows 11
- Visual Studio Code
- Internet access for the Marketplace install
- Python only if the fallback optimizer is used
The extension is light and fits normal desktop use.
If you use the Marketplace version:
- Open VS Code
- Open the Extensions panel
- Search for Prompt Optimizer Lite
- Install it
- Restart VS Code if needed
If you use the release download:
- Download the latest release from the Releases page
- Open VS Code
- Install the extension from the downloaded file if your release package includes one
- Reload VS Code
Use prompt-optimizer-lite when you want to:
- make a prompt easier to understand
- turn a vague request into a clear one
- prepare prompts for chat assistants
- keep a prompt short and focused
- save time rewriting the same request
It works well for plain tasks like:
- writing support requests
- drafting content prompts
- cleaning up instructions
- rewording notes into a better prompt
Before:
- make this better and more clear for the AI
After:
- Rewrite this prompt so it is clear, short, and specific
If the chat model does not respond, the extension tries the local Python optimizer. If that is not available, it uses a built-in template. This gives you a working result even when the first option fails.
- Open VS Code
- Open Chat or select text in the editor
- Run prompt optimization
- Review the improved prompt
- Copy it, insert it, or replace the original text
- write one prompt at a time
- keep the request direct
- include the goal first
- remove extra words before optimizing
- use the output as a starting point and edit it if needed
Depending on the release, you may see:
- a ZIP file
- a VS Code extension package
- a Windows installer
- support files for the local Python fallback
If you download a ZIP file, extract it before opening the contents
- restart VS Code
- check the Extensions panel again
- make sure the file finished downloading
- open Chat or select text first
- try the command again
- restart VS Code
- give the prompt more detail
- make the request more specific
- remove long background text before optimizing
- check whether Python is installed
- try again after restarting VS Code
- use the built-in template result if needed
- Open the Releases page
- Find the newest release
- Download the Windows package
- Open or extract the file
- Install or load it in VS Code
- Start using
@promptoptor the editor command