All notable changes to the "write-assist-ai" extension will be documented in this file.
--
- Option to set a
proxyURL
(baseURL
) to the OpenAI calls gpt-4o-mini
model to the selection dropdown- Option to set/remove the separator text around AI response
CodeActions
stopped showing if user didn't enter the API Key
gpt-4o
model, and made it the default model
- Language specific configuration capability for System Prompts, QuickFix Actions & RewriteOptions.
- OpenAI config changes take effect immediately without a reload.
- Move the action prompt to user message prefix of the OpenAI API call.
- A new command in the
Command Palette
to set/reset the OpenAI API Key - A common
system prompt
config for use with the LLM calls - Dynamically configurable
Quick Fix
andRewrite
actions - Use of
v1/chat/completions
endpoint ofOpenAI
- API response display issue while setting the API Key
- Use of legacy
v1/completions
endpoint ofOpenAI
- Configurable temperature and model (with custom option) setting
- Runtime prompt for getting the OpenAI API Key from the user
- Store the OpenAI API Keys in
secretStorage
for enhanced security
- Existing setting for OpenAI API Key, and move any saved keys to
secretStorage
- Support for TeX/LaTeX files.
- The issue of failing API calls
- The base OpenAI model from
text-davinci-003
togpt-3.5-turbo-instruct
- Better error messaging in case of API errors
Update the demo gif
- Support for plain text files
- One more rewrite tone, authoritative
- More actions
- Rephrase text
- Suggest headlines
- Summarize selection
- Expand selection (Make text verbose)
- Shorten selection (Make text concise)
Changing the API Key type to String
- Fixed the extension configuration key
- Update README and add extension demo gif
Update README
- Add the extension icon
- Update the changelog
- Remove the browser field from webpack
- Change webpack environment to node
- Initial release