A useful little ritual
Less guessing.
More understanding.
You don’t need another dashboard to manage. You need a few clear answers about the AI tools you use every day. Start with these three.
01 / DAILY CHECK-IN
Start with the big picture.
Open your dashboard, choose Today, and see recorded tokens and their estimated cost together. Switch to 7 days when you want a wider view.
Open today's overviewWhat did my coding tools use today?
- One view across supported Claude Code, Codex, and Gemini logs.
- Recorded usage stays separate from missing data.
- Choose a day to inspect the records behind its totals.
02 / DAILY CHECK-IN
Find what deserves a closer look.
Review your model breakdown, then open the matching records. Understand where an estimate comes from before deciding whether to change your workflow.
Explore your recordsWhich model accounts for the most usage?
- Compare recorded usage within the same date range.
- Inspect input, output, cache reads, and cache creation.
- Check the rate source instead of guessing what a number means.
03 / DAILY CHECK-IN
Build a habit of checking, not guessing.
Come back after a coding session to inspect newly recorded activity. Your daily check-in points to the latest day and models worth reviewing, using the data your connector actually reports.
Open live activityWhat changed after I worked differently?
- Your last dashboard tab and date range are remembered in this browser.
- Review cache usage alongside input and output tokens.
- See live activity while your local connector is available.
Useful numbers, with their limits in plain sight.
Costs are estimates based on recorded tokens and public rates, not your provider invoice. A missing model rate stays unpriced. Usage that never appears in your local logs is outside this view. Nothing here promises automatic savings or changes your provider settings.