Local usage ledger
See where agent work is happening without turning token volume into a score.
Generated output is an operational signal, not a productivity score.
Ranked by generated output
| Project | Generated output | Agent work | Scheduled | Root / child | Comparable cost | Coverage | Change | Open |
|---|
Operational clues
Explicit schedule identity only
Which schedules consume usage, how many days they ran, and whether their output was measurable.
| Schedule | Runs | Active days | Generated output | Coverage |
|---|
Accepted attribution only
A Feature appears only with one unambiguous accepted attribution head.
| Feature | Agent work | Generated output | Coverage |
|---|
Limits beside totals
Newest first
Open one item for its accounting facts. Prompts, session ids, and paths are never returned.
This browser has not been let in yet.
Ask to pair and a six-digit code appears on the Mac — it is deliberately not in the reply to this page. Only somebody at that machine can finish this, and that is the point rather than the hurdle.
Look at the Mac.
Clawdline is showing six digits for two minutes. Type them here.
Sign in with the password.
The one set in Clawdline on the Mac, under Settings → Remote. If none has been set, pairing is the way in.
Look at the Mac.
Open Clawdline on the Mac, go to Settings → Cloud, and give it the code below. It is only good for a few minutes and only once.
This browser’s fingerprint is — the Mac shows the same one.