Tracr

I’ve tried every time tracker. Toggl, Clockify, the one built into Notion, the one built into Jira. They all want you to work the way they think you work — projects with sub-projects, billable vs. non-billable, weekly summary emails.

Tracr is different because it has exactly the features I need and nothing else.

What it does

You run it from the terminal. You start a timer with a short description. You stop it. That’s the core loop. The data lives in a local SQLite file. There’s a log command that shows what you’ve been doing. There’s a summary command that groups by tag or project.

No cloud sync. No web dashboard. No accounts. A file on your disk that you own.

Why raw SQL

I started with an ORM — Drizzle, specifically. It’s a good ORM. But I kept fighting it whenever I needed a query that didn’t fit the happy path. Aggregations, window functions, custom time bucketing — all things SQLite does well and all things that become awkward through an abstraction layer.

Switched to raw SQL with a thin query helper. The queries are readable, the migrations are just .sql files, and I understand exactly what’s hitting the database.

Status

Active development. I use it daily. Occasionally I add something I’ve been missing. The schema has been stable for a few months now.