Vendor adapter coverage

After image copy, the importer attempts to emit a nova-ve template per priority-vendor image. Adapters live under backend/scripts/import_eveng/adapters/. The current supported set:

Vendor	Image kind	Adapter
Arista vEOS	QEMU	arista_veos.py
Cisco CSR1000v	QEMU	cisco_csr1000v.py
Cisco IOL	IOL	cisco_iol.py
Cisco IOSv L2	QEMU	cisco_iosv_l2.py
Cisco IOSv L3	QEMU	cisco_iosv_l3.py
Cisco Dynamips (c3725, c7200)	Dynamips	dynamips.py — Phase 1 ships c3725 + c7200; c1700/c2600/c2691/c3620/c3640/c3745 are matched but flagged needs-manual-review until their runtime path lands
Generic Linux	QEMU	generic_linux.py (catch-all for unmatched qemu images)
Juniper vMX	QEMU (paired VM)	juniper_vmx.py
Juniper vQFX	QEMU (paired VM)	juniper_vqfx.py
Juniper vSRX	QEMU	juniper_vsrx.py
MikroTik CHR	QEMU	mikrotik_chr.py
VyOS	QEMU	vyos.py

Anything outside this set still has its image copied — only the template layer is skipped. The image is available; you can hand-author a template JSON under /var/lib/nova-ve/templates/<vendor>/<key>.json and the node modal will pick it up.

Paired-VM templates

Juniper vMX and vQFX ship as paired VMs — a routing-engine (RE) image and a packet-forwarding (PFE) image that must be launched together with a private bridge between them. The dedicated adapters emit a multi-node template that the "Add node" picker presents as a single entry; instantiating it produces the RE + PFE pair correctly wired up. See #188 for the paired-template flow.

Adding a new adapter

Each adapter inherits from adapters/base.py and implements convert(eveng_template) -> NovaTemplate. To add a new vendor:

1. Drop a Python file at backend/scripts/import_eveng/adapters/<vendor>.py.
2. Implement convert() — return a NovaTemplate matching the result schema in template_schema.py.
3. Register the adapter in adapters/__init__.py.
4. Add a unit test under backend/tests/import_eveng/ exercising a representative EVE-NG fixture.
5. Re-run the importer on a host that has the EVE-NG source tree.

See Development → Backend for the broader Python module layout.