01 / THE WORKFLOW
Give each job its own place
Keep the customer, machine and reported problem attached to a case, even when several machines are on the bench.
For the repair bench
Put Mech on a supported software job while you handle physical repairs, customer conversations and the machines that need your hands. Come back to a checked result and a usable case record.
Windows and Linux supported. macOS technical preview.
A practical first step
See OpalSpan at work
Generated from the current shared Mech interface with fictional data and scripted responses. This is a simulation, not real device work or measured AI timing. Repair-shop generation, 13 October 2026. Other bench devices are left untouched.
Another path for the bench
Boot a locally prepared repair USB and OpalSpan can help gather drive evidence, keep the case current and show what needs your hands next.
I'll use the **already locally booted** Rescue environment for read-only storage evidence. I won't run repair scans or write to the disk.
Arguments
{
"operation": "Read-only storage health and filesystem inventory"
}Result
{
"nvme_health": "critical warning",
"media_errors": 184,
"filesystem": "not mounted",
"writes_attempted": false
}Repeated media errors make a failing drive likely, but not yet physically confirmed. Write-heavy repair could reduce recovery options.
Arguments
{
"case_id": "demo-case"
}Result
{
"saved": true,
"status": "Blocked",
"blocker": "No verified backup or approved recovery destination"
}No writes were attempted. The case records the critical storage evidence and the blocker.
Next: confirm backup status, agree a recovery destination and physically assess or replace the drive before repair continues.
Generated from the current shared Mech interface with fictional data and scripted responses. This is a simulation, not real device work or measured AI timing. Recovery generation, 13 October 2026. An enrolled bench PC is shown, not a Live USB boot, recovered files or a physical repair.