Dear all,
I am trying to add docs here:
But it gives error, that it couldnt index any page. I think that this is something related to the credentials. Is it possible to pass credentials in some way so that it can index private repos?
Hey, we don’t currently support private repos in Cursor, including passing in credential as part of the scraper yet!
For now, your best bet is to add a local copy of your docs to your codebase, maybe in .md files for example, and then @ the folder when needed.
Thanks.
I keep all my process documentation in the private repo’s wiki. I’d really like the AI to be able to read this.
My workaround to create @doc access, I’m downloading the pages into a local temp folder. I pretty much have to sync it manually.
An MCP server may be of use here, as you could make a relatively lightweight one that can access your private wiki, and Cursor can then query them as and when it needs to.
Hey Dan! Seeing as it’s been a year since your post, and there are loads of examples of people needing to provide context from private repos, see:
- Add GitHub repo reference to the “Add Context” feature
- Indexing Private Repos
- Tutorial: Adding full repo context, pdfs and other docs - #5 by dbsx
- How to reference external codebase/files + keep file always in context? - #3 by jimmyn
- Referencing a private repo
what are Cursor’s plans to add this capability without forcing users to clone the repo into Cursor?