Your own page: Me

Teams and members. How you use Claude, what it costs, and where the spend went. Only you and whoever you report to can see it.

Me is everyone's own page. It shows only your own data, whatever your role. Only you, and whoever you report to, can see it. No prompt or response text is ever collected.

Choose a range at the top: 7 days, 30 days or All time.

The figures

Status: Active, Idle 7 d+, or Never connected, with your last activity.
Active days, sessions and prompts.
Hands-on / Claude working: your keyboard time against the time Claude spent on tools and answers.
API-equivalent: what your usage would cost at API prices, and how many times your seat price that is.

If your computer isn't set up yet, a notice says nothing can arrive, with a link to finish setup.

Way of working

A card names your way of working, with why and what to try next:

Idle
Chat-first
Hands-on
Agent-first
Multi-agent

Where the spend went

Each dollar of Claude spend is sorted into one bucket (productive, inefficient or wasteful) with the pattern behind it and a fix:

Runaway agent loop: stop and re-plan when one prompt runs on and on.
Rejected edits: agree the approach first (plan mode) so edits aren't thrown away.
Retries and API errors: check the error, then retry once with less context.
Frontier model on short tasks: use a smaller model for quick questions.
Low cache use: keep one session per task instead of restarting.

Recent sessions

From your session-end hook: when each session closed and the folder it ran in.