agentleFS
itentialitential/infra-agent-skills/skills/compute-instance-provisioning/SKILL.md

compute-instance-provisioning

Skill1 starsChanged 32 days agoApache-2.0View on GitHub

Create, read, resize, snapshot, power-control, and decommission virtual machine instances on a supported hypervisor or cloud platform — currently Amazon EC2, Google Compute Engine, VMware vSphere, and Morpheus. Instance lifecycle only — not a general cloud or hypervisor administration skill. Assumes an authenticated client for the target platform's API is already available; this skill does not provision or authenticate it. Check references/INDEX.md for which platforms and operations are covered.

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 0 are distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

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

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

    Not for containers or Kubernetes workloads — see container-workload-deployment.

    line 17
  2. Very specific:

    Not for upgrading network-device operating systems — see software-upgrade-lifecycle.

    line 18
  3. Very specific:

    Not for allocating addresses or recording the instance in an inventory — see ipam-allocation, dcim-asset-management, and …

    line 19

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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

Compute Instance Provisioning

Manages the lifecycle of a virtual machine instance — create, read, resize, snapshot, power-control, decommission — on whichever platform the caller is targeting, following that platform's own documented procedure exactly.

When to use this skill

  • Creating, reading, resizing, snapshotting, powering on/off or restarting, or deleting a virtual machine instance on a supported hypervisor or cloud platform.
  • Not for containers or Kubernetes workloads — see container-workload-deployment.
  • Not for upgrading network-device operating systems — see software-upgrade-lifecycle.
  • Not for allocating addresses or recording the instance in an inventory — see ipam-allocation, dcim-asset-management, and cmdb-asset-management. Where creating an instance requires attaching an existing network, volume, or address, that attachment is part of the platform's own create procedure; creating those objects themselves is out of scope.

Universal concepts (shared virtual-instance vocabulary, not a wire standard)

These are claims expected to hold for any hypervisor or cloud platform, and were checked against each platform listed in references/INDEX.md. Where a platform differs, its own procedure wins.

  • An instance has a lifecycle state that decides which operations are allowed, and write calls are usually asynchronous. Read the current state before changing an instance, and confirm the target state after; an accepted request is not a finished one.
  • Deleting an instance is final, and what happens to its storage depends on the platform and its settings. Treat deletion as a confirmation point and tell the operator which storage will go with it.
  • Some changes, such as resizing, may require the instance to be powered off, where the platform says so; some platforms allow certain changes while it runs. Power-off interrupts the workload and can lose volatile state; confirm before doing it, and let the platform's own procedure say what else changes.
  • What "snapshot" means differs by platform (a point-in-time copy of the whole machine, of its disks, or an image to launch from), and not every platform can revert an instance to one, or offers a snapshot of an instance at all. Do not assume either exists.
  • Repeating a create request can create a second instance unless the platform offers an idempotency mechanism; use it where it exists.

Procedure

  1. Identify the target platform — from the caller's request or the platform argument. Check references/INDEX.md; if no matching row exists, stop: this skill doesn't cover it yet.
  2. Open references/<platform>/procedure.md and follow it exactly — it contains that platform's full operational procedure, error handling, patterns, known limitations, tools, and verification checklist.
  3. Report the outcome using that platform procedure's own status vocabulary — don't normalize it away.

Known limitations

  • Platform coverage is in references/INDEX.md; a platform not listed there needs a new reference file before this skill can be used for it.
  • Scope is the instance lifecycle only. Networks, storage systems, images as standalone catalog objects, identity, and other parts of a cloud or hypervisor are out of scope even where the same platform offers them.

Contributing to this skill

SKILL.md only changes for two reasons: a bug in the universal concepts/router logic above, or adding a pointer to a newly-supported platform. Platform-specific procedure changes — a new error code, a corrected field, a changed API path — belong in that platform's own references/<platform>/procedure.md, never here.

Tools

Each platform assumes its own authenticated client — see references/INDEX.md for which, and that platform's procedure.md for the specific operations.

Verification checklist

  • Target platform confirmed against references/INDEX.md before proceeding
  • That platform's own procedure.md was followed, including its own confirmation gates on launch, power-on, power-off, reboot, resize, an image that reboots the instance, and deletion
  • Outcome reported using the platform procedure's own status vocabulary

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.