Ivan Villa Well, in some cases, it makes sense to get work done in an AI chat interface, like a Claude project, where our team is doing research, collaborating, and really tweaking an output result with full context. Then, we could pass this ouput directly to a pre-trained Super Agent to complete an action, just like we'd do with a real team member.

For a simple example, let's say I'm working through a problem in Claude, then I realize my solution would make a great blog idea, so I call "Todd" our blog Super Agent via the MCP (maybe there's a super agent skill) to go create a blog for a particular project, which would be easy if it's already in a client project in Claude. Todd would find that project in ClickUp, find the recurring content task, and complete it in the way it's trained to do. Other team departments won't need access to my chat thread since they can communicate directly with Todd on that task for feedback.

This would save me time by eliminating the need to find and add all that context to the task manually.

Derek Cummings That's a great example! The only tricky part is that an agent might take a while to run depending on how intricately you built it, so it won't be able to respond instantly in the conversation.

I think we could at least return a link to the conversation immediately. That way, when you message the Todd agent, we can respond with: "Successfully messaged the Todd agent. To follow up or see the chat, visit clickup.com/[link-to-todd-chat]"

Ivan Villa Found myself looking for this functionality again today. It would be nice to bring some of my super agents or Brain GPT into my VS code environment or be able to be load into my website to handle various things like fixing bug or be our on site chat bubble and bring in the sales/support chats into clickup as tasks. We could do all sorts I have tons of ideas honesty.

Ivan Villa could it calculate that it will or wont take some time, and let the user know that it will notify them once it has completed?

We’re building an internal Brain at our company, and this capability would help us a lot. Non-technical people across different teams are building various Super Agents in ClickUp, Claude, ChatGPT, and on internal servers running local models. What we’re trying to do is use our internal Brain to connect all of these agents together across different workflows, but ClickUp basically doesn’t provide this capability for us.

For example, we currently have a Super Agent that’s specifically designed for social. We want that when someone in the company Brain wants to write a full campaign plan, they can call the social agent, receive the social output, and add it to the outputs from other agents so they can produce the complete plan.

Or imagine someone in our internal Brain wants to create a landing page, and based on our internal data an initial file gets triggered. In that case, we need our landing-page-builder Super Agent to activate, generate the output, and the person should be able to receive it directly from our Brain or other agent active to build landing on production!