[feature request]: Add caching for request results

[hgaogao]

• It is recommended to implement caching for request results, avoiding the need to send requests every time. This is particularly beneficial when dealing with a large number of tags, as it prevents fetching all tags with each grouping.
• Adopting this approach ensures that user actions are not overwritten, providing a safeguard for their interactions.
• Consider offering an option to re-sort all tags （Maybe not, because Ungroup All can already achieve the goal）

[MichaelYuhe]

I have no experience in the caching OpenAI results before, I'll try to learn and finish that.
If there can be an option to sort group it will be great.

[jt-wang]

@hgaogao @MichaelYuhe:

Cache is possible by

1. intercept the request
2. Check the body with cache

Some materials https://developer.chrome.com/docs/workbox/modules/workbox-routing and https://web.dev/learn/pwa/serving

Though I use OpenAI API every day to boost my workflow, I still receive only below $2 invoice every month. Cache sounds like more a good-to-have feature, but really interesting. Let me know if you want it so much and I can contribute.

[MichaelYuhe]

The best way to save credits: don't use gpt-4 :D
I think this feature needs more discussion since it may have much work to do, and maybe conflicted with #35? I have no idea if the cache ways are the same for different models or other classify solutions