I do propose integrating various thinking traces natively into Kagi as they do provide superior results.
https://github.com/chirag127/clear-thought-mcp-server
https://github.com/waldzellai/clearthought-onepointfive
https://github.com/waldzellai/waldzell-mcp/blob/main/servers/server-stochasticthinking
Users are using sequential thinking in various applications to provide a deeper reasoning process similar to Ki Assistant.
