Any updates on New Enhancements?

I will be a little surprised if there’s any attempt to allow it to take actions before Docebo’s Agent Hub gets released in Q3.  Totally agree though that if my Power User could use it to query anything in their realm of visibility that would be a killer use case.  Reporting is often difficult for people so natural language ways to get the data back out would be huge.

Hi ​@Mirza, thank you for both the question and the framing, and it's exactly the kind of signal that helps us prioritize.

You're right that the current beta is self-focused by design: we deliberately started from the learner perspective to validate the MCP foundation before expanding scope. But the admin and power user MCP is where we are currently focusing, and the topics you've outlined are central to our roadmap for the next phase.

We're targeting the end of beta and general availability with the summer release (July 2026). No intermediate releases are locked in yet, though we may ship incremental drops to gather more targeted feedback. In the coming days, I'll share a survey in the next days through the PDG specifically to collect and prioritize use cases, your input here will directly shape what we build next.

Hi ​@mzirnhelt, that's a sharp observation, and you're actually pointing at something we think about a lot internally.

MCP Server and AgentHub are intentionally designed as complementary features, not competing ones. The way we see it: MCP gives you a standards-based protocol that connects Docebo to any external AI assistant. AgentHub, on the other hand, is where you get customizations and actions that go beyond what a MCP can express. Our direction is to have agents built in AgentHub be able to leverage all MCP Server actions, plus more. So the practical answer is: the right choice depends on your use case, your workflow, and how much customization you need. We'll make sure this distinction is clear as both features mature toward general availability.

+1 From me for admin/poweruser functions in the MCP. I’ve resorted to building something custom with claude skills and agents and connecting via direct API calls for now.

The ability to send a statement like “Find all changes made to ILT session XYZ in the last 6 months” and the agent knows to go look at the audit log, filter to the correct event types and targets, and cycle through the date range (since Audit log only shows up 90 days of records at a time) and kick out a simple report in a few seconds has been a major efficiency boost!

I also use it to query course and session data. “How many active courses are in category XYZ”, or “How many future sessions of Course ABC are scheduled in 2026”

​@brandonbillings - what you have going there seems very very clever. Would love to pick your brain about it at your leisure.

​@dklinger I’m happy to share any intel I’ve collected since starting down this path. Feel free to DM me!