agentleFS
karmabkarmab/kcli/skills/kcli-provider-development/SKILL.md

kcli-provider-development

Skill658 starsChanged 13 days agoApache-2.0View on GitHub

Guides implementation of new virtualization providers for kcli. Use when adding support for a new cloud platform, hypervisor, or infrastructure provider.

We split this skill into the rules it gives an agent

2 rules. About 0 are common, 1 is distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Add import and instantiation in kvirt/config.py (around lines 102-220):

    line 134
  2. Distinctive:

    Not all providers need to implement every method.

    line 195

≈ “Common” is an estimate: at least 1 in 5 of the 308,232 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

kcli Provider Development

Provider Architecture

Providers are located in kvirt/providers/ as subdirectories. Each provider implements the interface defined in kvirt/providers/sampleprovider.py (class Kbase).

Required Implementation Steps

1. Create Provider Directory

kvirt/providers/yourprovider/
├── __init__.py      # Contains your Kyourprovider class
└── (optional helpers)

2. Implement the Provider Class

Your class must:

  • Set self.conn attribute in __init__ (set to None if backend unreachable)
  • Set self.debug from the debug parameter
  • Return standardized response dicts from methods

Return Value Pattern:

# Success
return {'result': 'success'}

# Failure
return {'result': 'failure', 'reason': "VM %s not found" % name}

3. Core Methods to Implement

VM Lifecycle (required):

  • create() - Create VM with all parameters (cpus, memory, disks, nets, etc.)
  • start(name) - Start VM
  • stop(name, soft=False) - Stop VM
  • restart(name) - Restart VM (default calls start)
  • delete(name, snapshots=False, keep_disks=False) - Delete VM
  • list() - Return list of [name, state, ip, source, plan, profile]
  • info(name, output, fields, values, vm, debug) - Return VM details dict
  • exists(name) - Check if VM exists
  • status(name) - Return VM status
  • ip(name) - Return IP string
  • clone(old, new, full=False, start=False) - Clone VM
  • export(name, image=None) - Export VM as image
  • console(name, tunnel, tunnelhost, tunnelport, tunneluser, web) - Graphical console
  • serialconsole(name, web) - Serial console

Storage:

  • create_pool(name, poolpath, pooltype, user, thinpool)
  • delete_pool(name, full)
  • list_pools() - Return list of pool names
  • get_pool_path(pool) - Get pool path
  • add_disk(name, size, pool, thin, image, shareable, existing, interface, novm, overrides)
  • delete_disk(name, diskname, pool, novm)
  • create_disk(name, size, pool, thin, image) - Create standalone disk
  • list_disks() - Return dict {'diskname': {'pool': poolname, 'path': name}}
  • disk_exists(pool, name) - Check if disk exists
  • detach_disks(name) - Detach all disks from VM

Networking:

  • create_network(name, cidr, dhcp, nat, domain, plan, overrides)
  • delete_network(name, cidr, force)
  • update_network(name, dhcp, nat, domain, plan, overrides)
  • list_networks() - Return dict of networks
  • info_network(name) - Get network info
  • net_exists(name) - Check if network exists
  • network_ports(name) - List ports on network
  • add_nic(name, network, model)
  • delete_nic(name, interface)
  • update_nic(name, index, network) - Update NIC

Subnets (cloud providers):

  • create_subnet(name, cidr, dhcp, nat, domain, plan, overrides)
  • delete_subnet(name, force)
  • update_subnet(name, overrides)
  • list_subnets() - Return subnet dict
  • info_subnet(name) - Get subnet info

Images:

  • volumes(iso=False, extended=False) - List available images
  • add_image(url, pool, short, cmds, name, size, convert)
  • delete_image(image, pool)

Snapshots:

  • create_snapshot(name, base) - Create snapshot
  • delete_snapshot(name, base) - Delete snapshot
  • list_snapshots(base) - List snapshots (returns list)
  • revert_snapshot(name, base) - Revert to snapshot

Update Operations:

  • update_metadata(name, metatype, metavalue, append)
  • update_memory(name, memory)
  • update_cpus(name, numcpus)
  • update_start(name, start) - Set autostart
  • update_information(name, information) - Update info metadata
  • update_iso(name, iso) - Change attached ISO
  • update_flavor(name, flavor) - Change VM flavor

Buckets (object storage):

  • create_bucket(bucket, public) - Create storage bucket
  • delete_bucket(bucket) - Delete bucket
  • list_buckets() - List all buckets
  • list_bucketfiles(bucket) - List files in bucket
  • upload_to_bucket(bucket, path, overrides, temp_url, public)
  • download_from_bucket(bucket, path)
  • delete_from_bucket(bucket, path)

Security Groups (cloud providers):

  • create_security_group(name, overrides)
  • delete_security_group(name)
  • update_security_group(name, overrides)
  • list_security_groups(network) - List security groups

DNS:

  • reserve_dns(name, nets, domain, ip, alias, force, primary, instanceid)
  • list_dns_zones() - List DNS zones
  • dnsinfo(name) - Return (dnsclient, domain) for VM

Other:

  • close() - Clean up connection
  • info_host() - Return host info dict
  • vm_ports(name) - List ports on VM
  • list_flavors() - Return [[name, numcpus, memory], ...] if platform supports flavors

4. Register Provider in config.py

Add import and instantiation in kvirt/config.py (around lines 102-220):

elif self.type == 'yourprovider':
    # Get provider-specific options
    option1 = options.get('option1')
    if option1 is None:
        error("Missing option1 in configuration. Leaving")
        sys.exit(1)
    try:
        from kvirt.providers.yourprovider import Kyourprovider
    except Exception as e:
        exception = e if debug else None
        dependency_error('yourprovider', exception)
    k = Kyourprovider(option1=option1, debug=debug)

5. Add Dependencies to setup.py

YOURPROVIDER = ['required-package1', 'required-package2']

# Add to extras_require dict:
extras_require={
    'yourprovider': YOURPROVIDER,
    # ...
}

# Add to ALL list if needed:
ALL = EXTRAS + AWS + ... + YOURPROVIDER

Info Method Structure

The info() method should build a dict with these keys:

  • name, autostart, plan, profile, image, ip, memory, numcpus, creationdate
  • nets: list of {'device': device, 'mac': mac, 'net': network, 'type': network_type}
  • disks: list of {'device': device, 'size': disksize, 'format': diskformat, 'type': drivertype, 'path': path}
  • snapshots: list of {'snapshot': snapshot, 'current': current}

Then call: common.print_info(yamlinfo, output=output, fields=fields, values=values)

Reference Implementations

Study these existing providers:

  • kvirt/providers/kvm/ - Libvirt (most complete reference, ~4300 lines)
  • kvirt/providers/aws/ - AWS cloud (~2200 lines)
  • kvirt/providers/gcp/ - Google Cloud Platform (~2100 lines)
  • kvirt/providers/kubevirt/ - KubeVirt on Kubernetes (~1900 lines)
  • kvirt/providers/vsphere/ - VMware vSphere (~1900 lines)
  • kvirt/providers/azure/ - Microsoft Azure (~1500 lines)
  • kvirt/providers/ibm/ - IBM Cloud (~1500 lines)
  • kvirt/providers/ovirt/ - oVirt/RHV (~1400 lines)
  • kvirt/providers/openstack/ - OpenStack (~1350 lines)
  • kvirt/providers/proxmox/ - Proxmox VE (~1200 lines)
  • kvirt/providers/hcloud/ - Hetzner Cloud (~550 lines)
  • kvirt/providers/web/ - Web-based provider (~530 lines)
  • kvirt/providers/fake/ - Minimal stub for testing (~20 lines)

Provider Complexity Guide

Not all providers need to implement every method. Focus on:

  1. Minimum viable: create, delete, list, info, start, stop, exists, status, ip
  2. Storage: add_disk, delete_disk, list_disks, create_pool, delete_pool, list_pools
  3. Networking: create_network, delete_network, list_networks, net_exists
  4. Images: volumes, add_image, delete_image
  5. Advanced: Snapshots, cloning, export, buckets, security groups (as needed)

Methods can return {'result': 'success'} with a print("not implemented") for optional features.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.