​
Beta 5 (2024-02-20) - 14 tickets

  • Allow the chat textarea to expand to multiple rows
  • Show view button button in the chat messages
  • Fix bug with the Agent is typing message dissapearing after viewing another thread
  • Show average wait time for the AI Developer in the chat
  • Create a page for domains and API endpoints
  • Show special tool messages in the chat
  • Add functionality to revert to a previous version of a site
  • Add Sentry to the workers
  • Create a run after the first message is added to the chat
  • Prepare getsite9000 for the OpenAI GPT store launch
  • Fix issues with Auth0 login and signup not working properly
  • Add Google Auth to Auth0
  • Preview a particular version of the site
  • Fix a bug with login not working on Safari

​
Beta 4 (2024-02-15) - 20 tickets

  • Create infrastructure for scheduled workers using Rancher
  • Fix a bug with the site not shown when generated for the first time
  • Fix a bug with roasts not being shown in the chat
  • Fix a bug with threads not updating messages properly after a new thread is started
  • Fix a bug that changed the subdomain between deployments
  • Create a run after a message is added to the chat
  • Call the runs endpoint instead of waiting for the result of message post
  • Implement the runs endpoints
  • Create services for all third party integrations in the API
  • Remove OpenAI calls from the controllers
  • Create a forked process worker when a new run is create
  • Replace draft UUID with site id
  • Functionality to delete a site
  • Show run duration in the chat window
  • Show multiple versions in the chat window top bar
  • Refresh JWT token when expired
  • Send all the code (HTML, CSS, and JS) in a single call to the API
  • Option to create multiple threads of a site
  • Update version when site is updated
  • Reload the preview iframe when a new version is created

​
Beta 3 (2024-02-10) - 5 tickets

  • Refactor the frontend to use ReactJS instead of jQuery
  • Create an onboarding spage for creating a new site
  • Try to use GPT-3 Turbo for the first few messages then switch to GPT-4 Turbo
  • Add a cursor to the messages endpoint for pagination
  • Inject code in the iframe for version preview or multiple choice

​
Beta 2 (2024-02-05) - 10 tickets

  • Fix thread files
  • Autoexpand chat textarea
  • Show a timer on the chat send button
  • Use all the messages from the conversation when generating a completion
  • Add images limitation to the pricing page
  • Limit images generation
  • Add user to request object in the API controllers
  • Store the users in the database
  • Add more restrictions to anon users
  • Implement a chat completion alternative as a fallback to the assistant API

​
Beta 1 (2024-02-01) - 14 tickets

  • Show better error message when rate limit of the API
  • Configure Sentry for the API and the chat app
  • Show directory with microsites on getsite.online
  • Only allow to upload files when the agent is not working
  • Implement try/catch for the API controllers
  • Fix issues with failed requests and runs getting stuck
  • Update the instructions and tools for the assistant each time the server starts
  • Create privacy policy and terms of service for the site
  • Store the name of the site in the database
  • Show the not found page if an html file is requests, otherwise send a 404 header
  • Add a short timeout for the transcribing and completion API calls
  • Create a metadata file in the R2 bucket
  • Add feature to speak with the agent (TTS)
  • Implement user roles