I spent some time testing Remnus with a small project to see how it actually works in practice. First, I created a few pages for project notes and requirements, then turned some of the tasks into a database. The table view was simple to work with, and switching the same data into a Kanban board was probably my favorite part. I could move tasks between statuses and still keep all the related information in the same place. I also tried the calendar view and added deadlines to a few tasks. Being able to switch between table, board and calendar without duplicating the data felt really convenient. The most interesting part for me was the MCP integration. I connected Remnus to an AI coding tool and tested whether the agent could actually read my project notes and update tasks instead of just using Remnus as a reference. It worked much closer to how I would expect an AI-native workspace to work — the agent could access the same pages and task data I was working with. I also liked that the AI's changes remain visible in the workspace rather than disappearing into some hidden AI memory. That makes it much easier to review what the agent actually did. There are still a few areas where I think the product could use more polish, especially around some workflows, but overall the concept is very solid. The combination of a normal project workspace with native MCP support makes Remnus much more interesting to me than a standard Notion-style productivity tool.

I tried the public demo, which needs no sign-up. It opens a ready-made workspace for a small Paint-clone project with a start page, dashboard, map and sprint board. I switched the sprint board between Board and Table views; the backlog cards show priority and category clearly and the page loaded fast. The best part is the agent audit log on the start page, which shows what the AI agent actually did and made the product easy to understand in about two minutes. Also good: an honest note that demo changes are temporary, a clear three-step setup, and a token-saving comparison with a 'how we measured' link. Weak spots: the pitch assumes you already use coding agents, so non-technical teams may feel lost, and the demo shows only one example project. Limits of this review: I did not create an account or connect a real agent, so I cannot judge setup reliability. Clear idea and a strong first impression.

Used it to update a DB automatically. Worked quite well, but it had a bit of a learning curve so a better tutorial is needed.

I spent about 10 minutes testing Remnus. My first impression was positive. The interface is clean and modern, with a dark theme that's easy on the eyes. I explored the dashboard and tried connecting a simple task to see how the MCP integration works. The setup was straightforward, and I appreciated that there were no errors during my session. The workspace feels organized, and I can see how this would be useful for developers juggling multiple AI tools. That said, as a first-time user, I wasn't immediately sure where to start. A short onboarding tour or a "getting started" checklist would help new users understand the value faster. Also, the empty state on some screens coulduse placeholder examples or suggested actions. Overall, Remnus looks promising for developers working with AI agents. Nice job.