AI Agent Options

[samschultzponsys]

Attempted Debugging

• I have read the debugging page

Searched GitHub Issues

• I have searched GitHub for the issue.

Describe the Scenario

What are the current recommended ways to have an AI agent that can be trained on Bookstack data, I believe there was something called Danswer that has since been rebranded, but just wanted to see if theres a go-to for the community.

Exact BookStack Version

Latest

Log Content

Hosting Environment

Docker

[ssddanbrown]

Hi @samschultzponsys,

Danswer (Now Onyx) is still probably your best bet for properly parsing BookStack content to be searchable via LLM.
I had been playing with something more native/built-in via #5552, but I'm not sure on the direction/implementation that may take.

As some alternative options, there are some MCP server projects out there like this one.

One of our sponsors, SiteSpeakAI, also provide services for this kind of thing: https://sitespeak.ai/bookstack