AI Super Agents

I find I can't trust the AI at all. It constantly misses tasks, especially recurring ones. If I have a task overdue and decide to bring it forward, it doesn't pick that up. It should see the history and be able to say what the last close off date was etc

In an effort to understand how I can add to an Agents knowledge, (I was asking it wether or not I coudl add to the doc myself, or if it would add for me. And if I could change the title of the doc. Which it told me I could because it refrences it in its own way, and the name change wont effect it. It seems to be ok because it still sees the doc as youll discover in the first image...It sees the text added which proves it still is refreencing the correct doc. It doesnt see the images added though. Please let me know if I'm doing something wrong, and or if this gets resolved. Hope this helps.

I love the simple conversational process to develop Super Agents, but once created, a user has to manually go through the instructions and create the needed adjustments that come from using and testing the agent in real life. It would be great if there was an additional tab within the Super Agent backend/setup area where I can conversationally provide feedback on the Super Agent's performance and any changes I would like to/need to see, and then the Agent automatically implements the needed adjustments to the existing instructions using the same professional, technical, detailed approach that it used when initially creating the instructions.

I’d love to see ClickUp support agents that can be assigned work just like users. The idea is to create agents with defined personas (name, role, decision style) and attach instructions or workflow automations to them. This would allow us to assign tasks directly to agents, who would then follow their logic or AI to complete the work, log actions, and collaborate with other agents or humans. Making agents true workflow participants (not just background automations) would add clarity, accountability, and flexibility—enabling scalable, transparent, and modular teamwork as our organization grows. This would be a game-changer for process automation and hybrid human-AI collaboration in ClickUp!