RAG -> Vector search -> means that your documents are not indexed as full text but as Vectorized objects which mean that then you can search using concepts instead of exacts strings you would use with a regular "Fulltext search".
This makes the search less precise and more powerful at the same time (ie it could look clever to some extent).
I've been a user of Day One for some time, but your proposition is very interesting nonetheless. I love the fact that the data is a file. I'm going to see if it's possible to migrate one way or another...
I started looking and the conversion is relatively straightforward (for a LLM), I'll try to write code with this and get back to you, if this gives something?
I never thought this was intentional too. I had tried to let the model show me its tool and it had refused. The article on the House MD role play was a nice way for me to try this out on Kagi. Peace out
Using the House MD. prompt extraction roleplay from https://hiddenlayer.com/innovation-hub/novel-universal-bypas..., it was possible to extract the Assistant prompt, not really interesting, but also the ResearchAgent's prompt and tools provided to the Assistants inside Kagi.
I'm discovering that Claude is included in Kagi Ultimate and this is also slightly blowing my mind... Probably going to do the switch.
The fact that it's expensive is true but for me largely compensated by the niceties of the service. I don't really like the fact that they use Yandex either but the other search engines are not really satisfying for me anymore.
I've used Kagi Pro for several months now and it's working great for my personal and professional needs. The only thing I'm missing is the shopping features but I can with that and switch back to Google when I'm looking to spend my money on physical goods...
https://janet-lang.org/