You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Jupyter-server uses jupyter_server.services.contents.manager.ContentsManager for interactions with the filesystem. This gives users the ability to attach external file systems to Jupyter, for example hosted on object storage (i.e. S3 or HDFS).
Now when using Jupyter-AI, the /generate and @file methods, will try to read from local filesystem. However, when an non-local Contents-Manager is used, the files never show up in the JupyterLab UI.
Reproduce
Not really relevant
Expected behavior
For the UI to show contents generated by Jupyter-AI commands, such as /export and /generate.
For the @file context provider to load files shown in the UI.
The text was updated successfully, but these errors were encountered:
@MaicoTimmerman Thank you for opening an issue for this. I agree this ought to be fixed; JupyterLab extensions should be expected to have strong integrations with the rest of the JupyterLab ecosystem.
I've added this to the Jupyter AI v3 milestone for us to track. I'll keep this in mind as I continue to iterate on the chat command design for v3.
Description
Jupyter-server uses
jupyter_server.services.contents.manager.ContentsManager
for interactions with the filesystem. This gives users the ability to attach external file systems to Jupyter, for example hosted on object storage (i.e. S3 or HDFS).Now when using Jupyter-AI, the
/generate
and@file
methods, will try to read from local filesystem. However, when an non-local Contents-Manager is used, the files never show up in the JupyterLab UI.Reproduce
Not really relevant
Expected behavior
/export
and/generate
.@file
context provider to load files shown in the UI.The text was updated successfully, but these errors were encountered: