Monitoring
2/4seats reporting
1open must-fix findings
2drift alerts
64%highest seat cost-cap usage
Fleet health
| Seat | Verdict | Last scan | Must-fix | Recommended | Cost cap |
|---|---|---|---|---|---|
| Front desk PC | 2 recommended | 2026-06-09 | 0 | 2 | 31% |
| Owner laptop | 1 must-fix | 2026-06-09 | 1 | 4 | 64% |
| Bookkeeper desktop | not reporting | 2026-05-22 | 0 | 1 | 8% |
| Marketing laptop | not enrolled | - | - | - | - |
Drift alerts
2026-06-09Owner laptop: Weekly scan exited 1: plaintext .env under ~/.claude (must-fix). Fix routes through clevskill-vault.
2026-06-05Bookkeeper desktop: No scan received for 14 days; seat marked stale. Machine off or telemetry disabled.
An alert fires when a weekly scan reports a must-fix finding or a seat stops reporting. Fixes route through consent-gated tools; nothing auto-applies.
Exclusions (findings you intentionally keep)
Sometimes a finding is on purpose: your setup needs it that way. Add part of the finding's title below and it stops counting against your health from each desktop's next scan. Exclusions never disappear silently: they are listed here and in your weekly health email.
No exclusions set.
Weekly digest
Every Monday, the account owner gets the "Your AI is healthy" email with this fleet scoreboard. It sends automatically while your subscription is active; no setup needed.