EXTENSIONS

exe.dev VM lifecycle model. Wraps the exe.dev HTTPS API to manage VMs as

Comprehensive libvirt/virsh management — VMs, networks, storage pools,

Firecracker microVM lifecycle management via SSH + the Unix-socket REST API.

Azure infrastructure management via az CLI — 43 model types covering compute, networking, data, security, RBAC, Azure Policy, Defender for Cloud, Entra directory, monitoring, DNS, DevOps, Azure AI Foundry (accounts, model deployments, projects, quota), AI Search, Cosmos DB, PostgreSQL Flexible Server, Static Web Apps, Service Bus, Event Grid, Recovery Services, Log Analytics, subscription-wide topology with Mermaid diagrams and cost estimation, actual-spend cost analysis and waste auditing via Cost Management/Resource Graph/Advisor, and the Azure AI Vision Face REST API for identity-aware room services.

Desired-state local libvirt VM pool reconciler for Swamp. Defines, starts, removes, and publishes per-VM connection facts for downstream SSH/config models.

Lifecycle management for a Colima VM (macOS container runtime, https://github.com/abiosoft/colima) via the local colima CLI: start and provision, stop, restart, delete, sync VM status, and exec commands inside the VM. One model instance targets one Colima profile.

Xen Orchestra (XO) / XCP-NG VM lifecycle model over the REST API (/rest/v0): create a VM from a template with cloud-init, start, stop (clean/hard), snapshot, and destroy, plus template/pool/SR/network discovery so UUIDs are resolved rather than hardcoded. Token auth via the `authenticationToken` cookie (vaulted); plain HTTPS through one injectable fetch seam, TLS verification always on (optional private-CA trust via caCert).

Convenience Arch Linux VM image preparer for Swamp: downloads the Arch cloud image, creates a qcow2 overlay, and writes a NoCloud seed ISO.

Generic qcow2 base image downloader and overlay disk preparer for Swamp.

Generic cloud-init NoCloud seed ISO generator for Swamp. Renders user-data, meta-data, and optional network-config into an ISO 9660 image with CIDATA volume label, optionally writing the ISO to a host filesystem path.

Full sceptre-phenix control via the HTTP API, split into six models — config, experiment, scorch, vm, cluster and user: config CRUD (Topology/Scenario/Experiment/Image/User), the experiment lifecycle (create/start/stop/delete, schedule, apps, topology, files, trigger), scorch run control (start, poll-to-completion, status, cancel, output tarballs), VM control (inventory, reconfigure, power, snapshots, captures, commit), and cluster/server reads (hosts, disks, apps, topologies, version, features, settings) plus identity (users, roles, long-lived tokens).

Full minimega control via its command socket (or `minimega -e`), split into eight models — vm, vmconfig, namespace, network, cc, disk, mesh and host: VM lifecycle (launch/start/stop/kill, info, tags, net, screenshots), the launch config template, namespaces and the scheduler, networking (bridges, taps, VLANs, QoS, captures, router, dnsmasq), command-and-control (miniccc), disk images, the mesh/deploy fabric, and host/system ops (host, check, version, files, vnc, logging) plus a generic command passthrough.