agentleFS
ai-workspace-labai-workspace-lab/xworkspace-core-skills/skills/operations-management/capacity-cost-and-resource-lifecycle/SKILL.md

capacity-cost-and-resource-lifecycle

Skill8 starsChanged 5 months agoMITView on GitHub

Assess capacity, cost, resource ownership, VPS resizing, replacement, snapshots, reserved IPs, retirement, and cleanup. Use before changing cloud plans, disks, instance types, autoscaling, reservations, tags, budgets, snapshots, or lifecycle state.

The file

We split this skill into the rules it gives an agent

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

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Treat a resource resize as a lifecycle change, not a simple plan edit.

    line 8
  2. Distinctive:

    Every billable resource needs owner, service/environment, provider/account, Terraform address/state key, instance/resource ID, tags, cost …

    line 12
  3. Distinctive:

    Before change, capture observed CPU, memory, disk used/free, IOPS, network, error/latency headroom, backup size, growth forecast, monthly …

    line 16

≈ “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

Capacity, Cost, and Resource Lifecycle

Treat a resource resize as a lifecycle change, not a simple plan edit. Provider capabilities, disk size, state ownership, data recovery, traffic routing, and cleanup are separate decisions.

1. Inventory and ownership

Every billable resource needs owner, service/environment, provider/account, Terraform address/state key, instance/resource ID, tags, cost center, creation reason, retention/expiry, backup class, and deletion authority. CMDB and Terraform are the runtime facts; spreadsheets or hand-maintained lists are not.

2. Capacity and cost preflight

Before change, capture observed CPU, memory, disk used/free, IOPS, network, error/latency headroom, backup size, growth forecast, monthly cost delta, and rollback capacity. Define the success condition and the duration of observation.

Provider plan changes must be classified before apply:

  • In-place supported upgrade: create a rollback point, apply, then verify capacity and service health.
  • Unsupported downgrade or disk shrink: create a replacement, restore application-level data, adopt Terraform state, regenerate CMDB/inventory, deploy, validate, then make an explicit traffic cutover.
  • Snapshot restore: verify source size, target capacity, OS/architecture compatibility, provider constraints, and retention. A snapshot larger than the target disk is a rejected plan, not a retry candidate.

3. Cost controls

Set budgets and anomaly thresholds by account/environment/service. Review orphaned volumes, snapshots, IPs, load balancers, images, and stopped instances on a defined cadence. Automation may report candidates, but deleting billable or recoverable resources requires exact enumeration, owner confirmation, and a rollback/retention decision.

4. Retirement

Retire in stages: remove traffic, verify no consumers, retain recovery point for the approved window, export final audit metadata, then delete the exact resource IDs. Terraform state removal, provider deletion, DNS release, snapshot cleanup, and credential revocation are separate checklist entries; authorization for one does not imply authorization for the others.

5. Feature environment leases and retirement

For feature-scoped preview environments, read Feature lane lifecycle. Track feature-level ownership, active UAT leases, isolated data and billable resources. One PR closing is not a retirement signal for the entire feature. Engineering candidate and acceptance gates belong to feature-lane-release-workflow.

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.