HN Explorer

Show HN: Copy Folder to ChatGPT in 0.6s – OpenRepoPrompt (github.com)

9 points by CompetitivePen ↗ HN
Over this past month I had the idea to build a 100% open-source MIT-licensed tool to simplify sharing code with LLMs, without the vendor lock-in you get from most SDKs. Right now, it’s way too hard to export your data or work freely with models like o1 PRO or Grok 3, especially since they don’t even have API access.

So I built OpenRepoPrompt, an open-source tool from wildberry-source that serializes files and folders into XML for LLMs. I coded/designed from 12PM -> 11PM on Saturday and 8AM -> 11PM on Sunday. There are still tons of features missing (I'm working on better file filtering next!) but I think it's reached that - buzzword alert! - MVP stage where I want to see if anyone's actually going to use the darn thing.

Here's the OpenPrompt repo: https://github.com/wildberry-source/open-repoprompt. Here are some screenshots I took while making it: https://github.com/wildberry-source/open-repoprompt/raw/main....

I think it has some potential. What do you guys think?

8 comments

[ 3.3 ms ] story [ 29.5 ms ] thread
siva7
One thing not widely known: Be careful when using the projects feature or the addon integration window (e.g. vscode) with chatgpt or Claude. They both use RAG under the hood to save on compute (without disclosing it), even if it would fit in the context window of the prompt. This usually results in worse answers.
benatkin
LLMs don't support zip files?

Here's the XML format: https://github.com/wildberry-source/open-repoprompt?tab=read...

Edit: after a web search, it seems not. One thing that would be nice in these projects would be to mention that it's just due to limitations, rather than suggest that the plain text is more natural to the LLMs. Also to the LLM and chat platforms, please don't add arbitrary restrictions to guide usage, if a zip file is too big reject it because it's too big, not because it's a zip.

elashri
I have some question that is relevant that the author or someone knowledgeable and have tried those might answer.

Does XML work better when feeding context to LLM or plain text or something like json? Did anyone try or have a theory about difference (as an input to the LLM not about processing the files/folders)?

Also wouldn't this be a good WASM usage or even just plain webpage? I know that processing large codebases like chrome/Linux would be better served via something else like Go program but would this even be useful? The context window of our best models is the real limit here and not the serialization speed of the context. using RAG is different league and out of consideration here of course.

I liked repoprompt [1] features during beta test but they then moved to $200 price tag which is hard for me to justify for my use case. Also it is Mac only and I mainly develop on Linux which is why I think something like webapp or WASM will be the more suitable option. I maybe wrong on several points here but no problem, I welcome any correction or explanation

[1] https://repoprompt.com

SparkyMcUnicorn
For the times I need this type of tool, I sometimes reach for repomix[0]. It does have a webpage[1], although I've never used it.

It supports XML, Markdown, and plain text. XML is the default, and I've found it to perform well, but it might be a good tool to test differences for yourself.

If you're looking for an alternative to repoprompt, I'd recommend checking out Aider's copy-paste mode[2].

[0] https://github.com/yamadashy/repomix

[1] https://repomix.com/

[2] https://aider.chat/docs/usage/copypaste.html

siva7
XML works better with Claude. I would assume the same for Openai.
KMnO4
I don’t think XML is necessary. Every one of those tags uses way more tokens than it needs to.
asadm
Is there a simple utility / script that takes gemini/grok/gpt output (usually markdown with changes, and apply it to the folder back.

I dont want to deal with "smartness" of aider. I WANT to use the chatgpt/ai-studio/grok UI for conversation but I also want to apply any changes it suggests back to my code easily.