issue: npm dependency swagger requires Internet to install

[userx8x8x8]

Check Existing Issues

• I have searched the existing issues and discussions.
• I am using the latest version of mcpo.

mcpo Version

0.0.19 python

Open WebUI Version (if applicable)

n/a

Operating System

ubuntu 24.04

Browser (if applicable)

Firefox

Confirmation

• I have read and followed all instructions in README.md.
• I am using the latest version of both MCPO and Open WebUI.
• I have included the browser console logs.
• I have included the Docker container logs.
• I have listed steps to reproduce the bug in detail.

Expected Behavior

Swagger UI boots when launching command mcpo --config /path/to/config.json

Actual Behavior

I am working on installing mcpo on a network that has strict firewall rules. The firewall rules are blocking access to the following website. https://cdn.jsdelivr.net/npm/swagger-ui-dist@5/swagger-us.css. I do however have a local npm repository setup that is hosting the swagger npm package at https://registry.npmjs.org/ I would like the npm dependency for the UI to come from an official npm source. Or you should bundle the dependencies with the python package or the docker install. I have tested both the pip and docker install on an online system without these firewall rules and I am able to successfully install it. I have also ran an npm install swagger and it worked but the mcpo package still reached out the cdn.jsdelivr.net site even though it is already downloaded.

Note that my config.json is a simple time server. The command is not the issue and I am able to successfully install the package without firewall rules.

Steps to Reproduce

mcpo --config /path/to/config.json

Logs & Screenshots

n/a

Additional Information

Note I don't really use github so sorry if this post is missing information.

[userx8x8x8]

A simple check if it can't download swagger from cdn.jsdelivr.net simply try the official registry.npmjs.org would fix this installation issue. You could also just bundle all the dependencies in the docker image. Or make an offline installer docker image. This is supposed to be Local AI.

[userx8x8x8]

Something like this https://pypi.org/project/fastapi-offline/