Contract
- Input: current cloud spend, workload inventory, performance requirements, budget constraints, team structure.
- Output: FinOps architecture + cost visibility design + rightsizing recommendations + budget and showback plan.
- Scope: designs FinOps strategy; does not execute infrastructure changes or purchases unless explicitly instructed.
- Rule: designs FinOps strategy; does not execute infrastructure changes or purchases unless explicitly instructed.
- Rule: documented standards override defaults; explicit project rules take precedence.
- Rule: if blocked by missing context or dependencies, surface the blocker before proceeding.
Provenance
| Question |
Answer |
| What is the source of truth? |
The user's request, originating spec/issue, and the skill's declared outputs |
| What is in scope? |
Work covered by the skill's acceptance criteria and completion rules |
| What is explicitly out of scope? |
Files, behaviors, and decisions outside the skill's declared boundary |
| Who or what consumes this artifact afterward? |
The next skill in the workflow or the user |
| What evidence proves it is done? |
Completion criteria met, artifact saved, validation passed |
| What risk remains? |
Subjective judgment calls, missing context, or external dependency failures |
Artifact
Emit CostOptimizationArtifact as both:
- JSON:
.agents/skills/platform/artifacts/cost-optimization/{identifier}.json
- Markdown view: same filename with
.md extension
Cost Optimization (FinOps)
Design a FinOps strategy for cloud cost visibility, rightsizing, spot instances, committed use, budgets, and showback/chargeback.
Process
1. Frame the cost optimisation problem
- Current spend baseline: total cloud spend, top cost drivers, spend by service, team, environment.
- Waste categories: idle resources, over-provisioned instances, unoptimised storage, data transfer costs, unused licenses.
- Performance requirements: minimum CPU, memory, IOPS, network throughput; SLA constraints.
- Team accountability: who owns which resources; showback vs chargeback model.
Completion criterion: spend baseline documented; waste categories identified; team accountability model defined.
2. Cost visibility and reporting
- Cost allocation: tags, labels, projects, folders; enforce tagging policy.
- Reporting cadence: daily, weekly, monthly; anomaly detection thresholds.
- Dashboards: spend by service, team, environment; forecast vs actual; unit economics (cost per request, cost per user).
- Alerting: budget threshold alerts, anomalous spend spikes, forecast exceedance.
Completion criterion: cost visibility design documented with allocation tags, reporting cadence, and alert thresholds.
3. Rightsizing and instance strategy
- Rightsizing analysis: CPU and memory utilisation metrics; instance type recommendations; storage tier selection.
- Spot and preemptible instances: suitable workloads (batch, stateless, fault-tolerant); checkpoint and retry strategy; fallback to on-demand.
- Autoscaling: vertical (instance type change) and horizontal (replica count); target tracking vs step scaling.
- Storage optimisation: lifecycle policies, tier selection (standard, infrequent, archive), snapshot retention.
Completion criterion: rightsizing recommendations documented with utilisation thresholds and instance strategy.
4. Committed use and reserved capacity
- Committed use discounts (CUD): 1-year or 3-year commitment; scope (resource, service, region); exchangeability.
- Savings plans: flexible across instance families, OS, tenancy; compute-only vs specific scopes.
- Reservation management: existing reservation coverage, expiry tracking, renewal process, reservation exchange.
- RI / CUD coverage targets: baseline load commitment; variable load on spot or on-demand.
Completion criterion: committed use strategy documented with coverage targets and renewal calendar.
5. Budgets, showback, and chargeback
- Budgets: per team, per environment, per project; alert thresholds (50%, 80%, 100%, 120%).
- Showback: cost visibility to teams without direct charge; transparency and accountability.
- Chargeback: direct cost allocation to teams or projects; reconciliation process; dispute handling.
- Guardrails: spend caps, approval workflow for large purchases, anomaly-based blocking.
Completion criterion: budget, alert, and showback/chargeback plan saved.
Rules
- Rule: enforce tagging and cost allocation policies before implementing visibility or chargeback.
- Rule: rightsizing recommendations must be validated against performance requirements and SLAs before execution.
- Rule: use spot instances only for fault-tolerant workloads with checkpoint and retry; never for stateful or latency-critical workloads without fallback.
- Rule: committed use discounts should cover baseline load; variable load should stay on spot or on-demand.
- Rule: alert on budget thresholds before overspend; implement approval workflow for large or anomalous purchases.
Completion
- the skill's completion criteria are explicitly checked
- the artifact is saved and validated
- any blockers or skipped validations are documented
- the next consumer is identified or the work is handed off
@include .agents/skills/platform/contract-base.xml
1---2name: "cost-optimization"3category: "devops"4maturity: "stable"5version: "1"6description: "Design FinOps strategies — rightsizing, spot instances, committed use, budgets, showback/chargeback — with cost visibility and optimisation guardrails."7license: "MIT"8capabilities:9 - apply cost optimization methods within explicit devops boundaries10 - analyze design finops strategies — rightsizing, spot instances, committed use, budgets, showback/chargeback — with cost visibility and optimisation guardrails with assumptions, constraints, and failure modes11 - produce and validate a devops artifact with evidence, decisions, and next steps12outputs: "FinOps architecture + cost visibility design + rightsizing recommendations + budget and showback plan."13sideEffects: []14dependencies: []15stopCondition: "Architecture diagram saved; visibility design documented; rightsizing recommendations complete; budget and showback plan filled."16risk: "medium"17trustTier: "3"18maxIterations: "7"19promptVersion: "2.0"20artifactType: "devops"21modelTier: "reasoning"22evaluators: ["behavioral", "regression"]23fixturesPath: .agents/skills/platform/fixtures/behavioral/cost-optimization.json24diataxis: "how-to"25tags: ["devops"]26compatibility: []27approvalRequired: "true"28approvalFor: []29metadata:30 source: https://skillmd.com/skills/quantumquirkxyz/cost-optimization31---3233## Contract3435- Input: current cloud spend, workload inventory, performance requirements, budget constraints, team structure.36- Output: FinOps architecture + cost visibility design + rightsizing recommendations + budget and showback plan.37- Scope: designs FinOps strategy; does not execute infrastructure changes or purchases unless explicitly instructed.38- Rule: designs FinOps strategy; does not execute infrastructure changes or purchases unless explicitly instructed.39- Rule: documented standards override defaults; explicit project rules take precedence.40- Rule: if blocked by missing context or dependencies, surface the blocker before proceeding.4142## Provenance4344| Question | Answer |45|---|---|46| What is the source of truth? | The user's request, originating spec/issue, and the skill's declared outputs |47| What is in scope? | Work covered by the skill's acceptance criteria and completion rules |48| What is explicitly out of scope? | Files, behaviors, and decisions outside the skill's declared boundary |49| Who or what consumes this artifact afterward? | The next skill in the workflow or the user |50| What evidence proves it is done? | Completion criteria met, artifact saved, validation passed |51| What risk remains? | Subjective judgment calls, missing context, or external dependency failures525354## Artifact5556Emit `CostOptimizationArtifact` as both:57- JSON: `.agents/skills/platform/artifacts/cost-optimization/{identifier}.json`58- Markdown view: same filename with `.md` extension596061# 62# 63# Cost Optimization (FinOps)6465Design a **FinOps** strategy for cloud cost visibility, rightsizing, spot instances, committed use, budgets, and showback/chargeback.6667## Process6869### 1. Frame the cost optimisation problem70- Current spend baseline: total cloud spend, top cost drivers, spend by service, team, environment.71- Waste categories: idle resources, over-provisioned instances, unoptimised storage, data transfer costs, unused licenses.72- Performance requirements: minimum CPU, memory, IOPS, network throughput; SLA constraints.73- Team accountability: who owns which resources; showback vs chargeback model.7475**Completion criterion:** spend baseline documented; waste categories identified; team accountability model defined.7677### 2. Cost visibility and reporting78- Cost allocation: tags, labels, projects, folders; enforce tagging policy.79- Reporting cadence: daily, weekly, monthly; anomaly detection thresholds.80- Dashboards: spend by service, team, environment; forecast vs actual; unit economics (cost per request, cost per user).81- Alerting: budget threshold alerts, anomalous spend spikes, forecast exceedance.8283**Completion criterion:** cost visibility design documented with allocation tags, reporting cadence, and alert thresholds.8485### 3. Rightsizing and instance strategy86- Rightsizing analysis: CPU and memory utilisation metrics; instance type recommendations; storage tier selection.87- Spot and preemptible instances: suitable workloads (batch, stateless, fault-tolerant); checkpoint and retry strategy; fallback to on-demand.88- Autoscaling: vertical (instance type change) and horizontal (replica count); target tracking vs step scaling.89- Storage optimisation: lifecycle policies, tier selection (standard, infrequent, archive), snapshot retention.9091**Completion criterion:** rightsizing recommendations documented with utilisation thresholds and instance strategy.9293### 4. Committed use and reserved capacity94- Committed use discounts (CUD): 1-year or 3-year commitment; scope (resource, service, region); exchangeability.95- Savings plans: flexible across instance families, OS, tenancy; compute-only vs specific scopes.96- Reservation management: existing reservation coverage, expiry tracking, renewal process, reservation exchange.97- RI / CUD coverage targets: baseline load commitment; variable load on spot or on-demand.9899**Completion criterion:** committed use strategy documented with coverage targets and renewal calendar.100101### 5. Budgets, showback, and chargeback102- Budgets: per team, per environment, per project; alert thresholds (50%, 80%, 100%, 120%).103- Showback: cost visibility to teams without direct charge; transparency and accountability.104- Chargeback: direct cost allocation to teams or projects; reconciliation process; dispute handling.105- Guardrails: spend caps, approval workflow for large purchases, anomaly-based blocking.106107**Completion criterion:** budget, alert, and showback/chargeback plan saved.108109## Rules110111- Rule: enforce tagging and cost allocation policies before implementing visibility or chargeback.112- Rule: rightsizing recommendations must be validated against performance requirements and SLAs before execution.113- Rule: use spot instances only for fault-tolerant workloads with checkpoint and retry; never for stateful or latency-critical workloads without fallback.114- Rule: committed use discounts should cover baseline load; variable load should stay on spot or on-demand.115- Rule: alert on budget thresholds before overspend; implement approval workflow for large or anomalous purchases.116117## Completion118119- the skill's completion criteria are explicitly checked120- the artifact is saved and validated121- any blockers or skipped validations are documented122- the next consumer is identified or the work is handed off123---124@include .agents/skills/platform/contract-base.xml