How I Stopped Things From Falling Through the Cracks

As long as I have been working, task management always felt like overhead. Time that could have gone into actual work, spent on organising it instead.

Our company moves fast and has no strict processes unless you make your own. That works fine when you are by yourself. In a team of more than three, it breaks down quickly.

I tried lists, task management tools, all of it. They always got left behind the moment things got urgent. Every system I tried added more friction than it saved, and the context switching between managing work and doing work drained more energy than I expected.

So I added an agent to the mix. Not to replace a system — I did not want another system. The idea was simple: put in the least effort possible while still keeping things on track, dependencies managed, and work assignments clear. I told the agent to own all of that and be strict about it.

It did the data entry itself. I just talked about what needed to happen and it would arrange everything. Then it would feed tasks back to me one at a time, dependencies in order, nothing out of sequence.

It did not feel like using a tool. It felt like talking to a colleague who genuinely cared about how much I was taking on and whether I could actually finish it. It would push back when I was overloading a day. It would flag when something I wanted to assign was not ready to be assigned yet.

It also became the place where I think through new features before they go to the team. I could have the full design conversation without touching the codebase — just working through what the feature needed to do, what depended on what, what questions needed answering first. That output would then go to a coding agent that would pick it up in the context of the actual system and build it.

I do what I am good at. The agent handles the rest without burning through my energy in the process.