admineral / openai-assistant-api-chat Goto Github PK

Hello, here is a repo with a number of improvements, including image display, file downloads, loading assistants, communication history, and the ability to share assistants and conversations.

https://github.com/Direct-Technologies-CZ/OpenAI-Assistant-API-Chat

feel free to merge it.

i have an assistant ID than i added in the .env and when i ask assistant "who are you" it is correct. But When i start/load the promp it always state in the terminal text: { value: 'You are a Pirate! introduce yourself', annotations: [] } i have looked at OpenAI, Assistant and what i can see/know i have nothing naming a pirate. Is there something im missing or how can i remove this and make the prompt act like it should when loading?

Please add a possibility to upload a file into the running conversation, not just at the moment the assistant is initialized.

Not an issue, just don't know where to ask. I really could use the TTS from eleven labs and Whisper put in this, dunno if you'd be willing to add it for a fee or something but we have some folks here in that don't type too well and it would sure be cool. Please shoot me an email and let me know [email protected] I really hope it's possible cuz the last proton in the universe would decay before I could even think of being able to do that lol. Sorry for the intrusion.

Currently this works amazingly! However, I'd like to be able to show a list of existing assistants that have been created already and then allow someone to choose which assistant to chat with. Would this be something that could be done easily? I don't know node.js very well, I'm normally a PHP HTML CSS developer lol

Regarding OpenAI Assistants creation, should we split them with different permission?

1. Admin interface: allow to create assistants, upload files, or even define system prompts, of course, can chat as well.
2. regular users: Allow to chats.

Would you consider changing the chat introduction from Pirate to Professional or adding a choice for the user to specify as a parameter?

got this error

472-dcc473fb6ec49a48.js:1 Error uploading file: Error: File ID is undefined

Great repo, thanks! Do you take on private clients to implement the above and other projects?

I am interested in help using this repo.

plz check your vercel demo site.
I can't send any message.
https://open-ai-assistant-api-chat.vercel.app/

Hi when I deploy to vercel with my OPENAI_API_KEY and ASSISTANT_ID I got the error "Failed to run assistant"

I see runAssistant() just post a request but how do I fix this error ?

  export const runAssistant = async (assistantId, threadId) => {
    console.log('Running assistant...');
    const response = await fetch('/api/runAssistant', {
      method: 'POST',
      headers: { 'Content-Type': 'application/json' },
      body: JSON.stringify({ assistantId, threadId }),
    });
    if (!response.ok) {
      console.error('Failed to run assistant');
      throw new Error('Failed to run assistant');
    }

From the console I saw Thread ID: undefined, maybe this is the reason.

Chat thread created successfully. Thread ID: undefined

这个shish是什么问题

How to create multiple assistants

Would love to see a CLI version for this, im not really in to the upload file, just send a question with a specific Assistan role. but to not use a webbrowser and use the CLI instead would be perfect if possible :)

How should I go about bypassing this screen so it defaults to a new chat window?

Thanks

Hi just wondering how the size, and type of files uploaded affect the performance of the assistants API.