Overview
Overview is the dashboard home — the screen you land on when you open the dashboard, a glance at how your recent AI-assisted coding is going over the date range you pick in the top-right. Its subtitle sums it up: vibe coding performance at a glance.
The four tiles
Section titled “The four tiles”Across the top, each tile shows a headline number for the range; most also show the change from the previous period of the same length:
- Prompt Score — your average prompt grade, 0–100 with a letter grade. It’s the number to watch; a rising trend means your prompting is getting sharper.
- Total Tokens — the tokens you used in the range.
- Total Cost — the API-equivalent cost of that usage (no change chip).
- Sessions — how many Claude Code sessions you ran.
The change chip beside most of them shows the direction at a glance, so a heavy period reads differently from a light one before you look at any chart.
Prompt Score Trend
Section titled “Prompt Score Trend”A bar chart of your Prompt Score across the range — one bar per period, colored by grade band — so you can see whether your score is climbing, holding, or slipping.
Token Usage and Cost Usage
Section titled “Token Usage and Cost Usage”Two charts side by side. Token Usage breaks your tokens into Input, Output, Cache Read, and Cache Write. Cost Usage shows the API cost for the range. Together they show where your usage and spend are going.
Daily Activity
Section titled “Daily Activity”A calendar heatmap of how intense each day was. Switch the metric between Requests, Tokens, and Cost, and page through it month by month — the darker a day, the busier it was, so you can spot when you lean on AI hardest and when you don’t.