VM verification — Kali (campaign 1)¶
Second target of the VM ladder (docs/contributing/vm-testing.md); the
Parrot record explains the method. Kali maps onto the kali-rolling CI
target, so these are pass/fail results against current claims.
Date: 2026-08-29
Engine: commit 4020d73
Guest: Kali GNU/Linux Rolling (ID=kali, VERSION_ID=2026.3), x86_64,
Python 3.14.6, fully updated at baseline
Host: KVM/QEMU domain Kali_Dev_Box, qcow2, NAT
Snapshots: cold clean-baseline and cold station-installed, both
taken with the six-item prep checklist applied (the maintainer had applied
the NOPASSWD drop-in and enabled sshd before handoff; sleep-mask, final
updates and the cold shutdown were done over SSH)
Results¶
| Test | Result |
|---|---|
hammunition status |
Detects ID=kali, version=2026.3, arch=x86_64, Debian family yes. 225 of 225 manifests resolve — the two Kali-gated blocks (arduino-cli, soapysdr-module-plutosdr) that are deliberate gaps on Parrot light up here, exactly as the capability matrix claims. |
install station --yes --refresh |
Both commands completed and confirmed; all 10 packages independently verified ii. |
| Idempotency — same command again | Nothing to do. |
uninstall twclock --yes |
Completed and confirmed; dpkg shows rc afterwards. Reinstalled before the checkpoint. |
install sdrpp --dry-run |
The Kali-only manifest resolves to a one-command apt install — the first target where sdrpp arrives by apt rather than a source build. |
No defects found. The engine work hardened on Parrot (uninstall, staged root configs, the station prompt) transferred without a change.
Not yet run¶
GUI launch checks, hardware passthrough, the linbpq configuration build (exercised on Parrot; repeating it per-target is M5 work), and the M5 re-verification list.