How do I export chat with AI?

Feature Requests
1

How do I export or share a chat with AI?
I want to be able to share my chat with AI same as ChatGPT does in order to share it with my coworkers.
Or at least be able to export it. Currently I have to copy/paste every chat message manually and it becomes very tedious work.
Screenshot is not an option, as I need to be able to search through it, and copy/paste content.

Are there any suggestions on how can I share/export quickly chat on Cursor AI?

16 Likes
2

No way right now, but noted as a feature request!

10 Likes
3

I’d like to endorse this feature request. Or at least, some way to name and preserve chats. They can be invaluable reference.

I’ve recently created a way to re-view ChatGPT conversations. Assuming you’re able to export using their default format easily, you could use it: GitHub - dreki/convo-formatter

5 Likes
4

It’s great to hear that this has been noted as a feature request. I’ll be eagerly awaiting its release!

2 Likes
5

+1 … we are using Cursor for teaching and it would greatly help to be able to share chats as examples or for feedback / debugging. thanks!

3 Likes
6

+1

Would like to store some really helpful conversations to Obsidian (or just standalone .MD file). Copy & Paste is clunky, and fomatting gets lost (code blocks especially).

2 Likes
7

I think if each pull request also had the conversation that were asked of the AI saved and checked in with the PR. It would help fine tune chain of thought for JIRA to PR.

I made a prototype that used the custom editor api to edit chats in vs code. But I want to be able to hook it into cursor chat.

8

+1
Hope this feature can release soon. It would be greatly appreciated

3 Likes
9

Here’s a quick workaround till this feature is implemented:

6 Likes
10

Just hit this stone wall using Cursor after just about switching fully over to using it! I pray this is implemented, as the key distinguishing feature of Cursor over VS Code is precisely the chat window - and if this chat window is lost and cannot be saved - my workflow quickly tops out and becomes debilitated for serious work.

Seems a critical feature I assumed would be top-of-list and already implemented.

Great program! This may be the show-stopper for me, however, in terms of using it.

4 Likes
11

Not sure if there’s a better way to do this, but +1 on this :handshake:

3 Likes
12

Please could we have an update/roadmap on this feature?

2 Likes
13

I consider this a very useful feature. I recently had a very long interaction with my friend Sonnet, and I would love to export it so I can format it, remove all the unnecessary parts, and save it as a reference. This would be helpful for sharing with my team, save it as a doc in the repo, and for using it to feed the LLM tomorrow.

3 Likes
14
  • 1 for this. Very useful for iterating on system design.
2 Likes
15

In case it assists anyone, I noticed there is a ‘Copy’ feature for an individual chat response, and documented it in this post, which produces markdown of the individual chat response, but not the entire chat. In that post, I included a request to Cursor to please add this same functionality for the entire chat.

1 Like
17

I use this little hack:

Have it generate a response with whatever model you want.

Use GPT-4o-mini

make a quick javascript website that puts the entire previous response in clipboard when you click a button

Then copy the html document with the copy button and paste to a new file with a html extension.

Open the html document in your browser and click the button.

The full output of the response is now in your clipboard.

I had used this for some create formatted MD files and most of the response being interpreted by chat.

It probably will not work for every scenario.

1 Like
18

I would also like to save a full chat, which is sometimes quite long. Could somebody please explain why this features is absent from almost all applications. Are there subtleties involved that make this difficult to implement? That is hard to believe since the last context saved in a multi-turn conversation is the full conversation, so it could be saved in a temporary buffer to be returned to the user upon demand. Could somebody please explain the difficulty to me? Thanks.

19

This will help me as I often want to do a chat to handoff tasks to another engineer to implement

20

I would also like to add that similar tools like Cody support this. Cody on VSCode will export all conversations as a json whereas Cody on JetBrains/Android Studio allow you to right click each history and export as JSON for the selected conversation.

1 Like
21

+1 this seems to me part of the basic functionality…