Hey everyone, I’m Karthik. I’ve been building a planner around a habit I haven’t quite managed to fix: assuming a task will take 30 minutes, spending an hour on it, then making the same estimate tomorrow.
I wanted to keep enough of the day to learn something from it before planning the next one.
The app is called Personal Planner. You place tasks on a timeline, start a timer while you work, and compare the time you planned with the time you actually spent.
The part I care most about is looking back. Daily and weekly reviews put your tasks, recorded time and notes together. If something didn’t happen, you can write why, so you have some context beyond another unfinished checkbox.
A few other things it supports:
- Day and week views with draggable time blocks.
- Recurring tasks and templates for regular plans.
- An inbox for things you haven’t scheduled yet.
- Experiments for following a routine over a few weeks.
It works offline without an account and stores your data on your device. It’s free and open source under the MIT license, with no subscriptions or paid feature tiers.
Android downloads are APKs from GitHub Releases; it isn’t on Google Play. There are Windows, Linux and macOS builds too. Optional sync uses your own Supabase project and needs setup.
Here is the website where you can find more details about the app: https://karthikshabari.github.io/PersonalPlanner/
Source code: https://github.com/Karthikshabari/PersonalPlanner
This is my first independently released app, and I use it every day myself. I’d especially appreciate feedback on how planning and reviewing feel on a phone.
When a task takes twice as long as expected, do you adjust your next estimate, or does tomorrow somehow get the same optimistic plan?
I reckon the offline no-account approach is right. Storing on device and making sync your own Supabase project is more friction than I’d take on. I think it’s the right trade. The reviews are only worth anything if the data stays put long enough to learn from, and asking people to trust a server with their honest time logs in week one is a lot. Good luck with the F-Droid queue.
2 questions, any ai disclosure? And any plan to submit this to fdroid for distribution?
Ai is used around 50-55% to genrate all the boilerplate and repetitive stuffs the core feature I implemented and I am trying to submit in fdroid but the review is very long that’s what holding me back :(
If you any doubts let me know and one more thing I want to understand how you guys are planning in general as well:)?


