Currently the Wiki pages are added to memory by sending Bro pdf documents for reading and remembering. This is because some pages aren't publicly available for Bro to use web search tool for reading. By using the wiki.js api, Bro could access the private pages by an API token. This will save a lot of openAI tokens to read the pdf documents.