BR Tasks
With thousands of task apps available, is it worth building your own? I wanted to see what I could gain from a tool that fits my work and my agents.
A task system of my own
With off-the-shelf apps, I kept fitting my habits to someone else’s design. I wanted to decide how I capture tasks, what priority means and how an agent helps. Today’s cost of building software makes that experiment worthwhile for me.
My role
I design the product and use it every day. I set the rules, connect other tools and make changes with coding agents. Daily use quickly reveals which ideas help and which make the task list harder to use.
Agents have their own way in
There is an Inbox, projects, priorities, a measure of task difficulty and a Focus view. REST and MCP let agents work on tasks alongside me. Project context lives separately in BR Projects. My phone, desktop app and agent can all use the same task system.
JEV understands what I am typing
In the experimental Inbox, I type a sentence and see the project, priority and other fields fill in as I go. I can adjust them before saving.
JEV is a TypeSafe model for small, specific decisions. It receives context and possible answers, selects the relevant ones and returns probabilities. In BR Tasks it identifies what parts of the sentence mean; application code calculates dates and durations.
I like how AI fits into an ordinary interaction here. I type a task and watch the form take shape. I enable it per device because I am still testing the experiment.
Result
I use it daily. When my needs change, I can change how the app works.
A private system for my own use. The presentation below uses example tasks.
