Przejdź do treści
Biblioteka skilli
DevOps i chmura · Analiza

enterprise-finops-cloud-cost

/enterprise-finops-cloud-cost

Analizuje koszty chmury AWS, GCP i Azure: alokację, rightsizing, rezerwacje i transfer danych.

Repozytorium
agent-skills ↗
Licencja
MIT
Ostatnia zmiana repozytorium
10.11.2026

Ocena i bezpieczeństwo

95/ 100Klasa: dobry
Format20 / 25
Frontmatter z nazwą i opisem, poprawna nazwa, opis do 1024 znaków, nazwa zgodna z folderem.
Odkrywalność25 / 25
Opis mówi, kiedy użyć skilla, ma sensowną długość i jest w 3. osobie.
Treść35 / 35
Co najmniej 150 słów i do 500 linii, nagłówki, listy kroków, przykłady albo kod.
Higiena15 / 15
Bez zaślepek typu TODO, bez ścieżek Windows, z licencją.
Ryzykobrak sygnałów w skanie
Pliki w paczce
1
Skrypty
0
Komentarze HTML
0
Adresy URL w SKILL.md
0
Przestarzałość

Brak sygnałów. Repozytorium zmienione 10.11.2026, bez odwołań do wycofanych modeli i API.

Licencja

MIT, otwarta: można używać i zmieniać, z zachowaniem informacji o autorze.

Co robi i kiedy

Analizuje koszty chmury AWS, GCP i Azure: alokację, rightsizing, rezerwacje i transfer danych.

Wyzwalacze z opisu

cloud bill too highfinopsright-sizingspot instancessavings plansreserved instancesegress costsunit economics cloud
Oryginalny opis (EN)

Use this skill when the user says 'cloud bill too high', 'finops', 'right-sizing', 'spot instances', 'savings plans', 'reserved instances', 'egress costs', 'unit economics cloud', 'cost allocation tags', 'showback', 'chargeback'. It applies the FinOps lifecycle (Inform, Optimize, Operate) to AWS/GCP/Azure spend: allocation, right-sizing, elasticity, commitment coverage and data-transfer architecture. Do NOT use for: Kubernetes autoscaler configuration (use devops/kubernetes-autoscaling), enterprise budget policy and approval governance (use enterprise/cost-governance), generic devops cost tool

Jak działa

  1. Purpose
  2. Agent Protocol
  3. FinOps Lifecycle
  4. Optimization Levers (do in this order)
  5. Data Transfer (often the hidden line item)
  6. Allocation & Accountability
  7. Unit Economics
  8. Anomaly Detection
  9. Anti-Patterns
  10. Checklist
Sekcje z nagłówków SKILL.md, w oryginale · 99 linii, 776 słów
Podgląd SKILL.md
---
name: enterprise-finops-cloud-cost
description: >
  Use this skill when the user says 'cloud bill too high', 'finops', 'right-sizing', 'spot instances', 'savings plans', 'reserved instances', 'egress costs', 'unit economics cloud', 'cost allocation tags', 'showback', 'chargeback'. It applies the FinOps lifecycle (Inform, Optimize, Operate) to AWS/GCP/Azure spend: allocation, right-sizing, elasticity, commitment coverage and data-transfer architecture. Do NOT use for: Kubernetes autoscaler configuration (use devops/kubernetes-autoscaling), enterprise budget policy and approval governance (use enterprise/cost-governance), generic devops cost tooling (use devops-finops).
version: "2.0.0"
author: "j4flmao"
license: "MIT"
tags: [enterprise, finops, cloud-cost, aws, gcp, azure, optimization]
---

# FinOps & Cloud Cost Optimization

## Purpose
Treat cloud cost as an engineering metric with an owner. Make spend visible per team/product, remove waste, buy commitments only for proven baseline load, and design architectures whose cost scales with business value (cost per order, per tenant, per request).

## Agent Protocol

### Trigger
"finops", "cloud cost", "aws bill", "right-sizing", "spot", "savings plan", "reserved instance", "committed use discount", "egress", "data transfer cost", "showback", "chargeback", "unit cost".

### Input Context
- Cloud provider(s), monthly spend, top 5 services by cost.
- Tagging/labeling coverage; account/project structure.
- Workload types: steady 24/7, business-hours, batch, interruptible.

### Output Artifact
Savings plan: ranked actions with estimated monthly savings, risk and owner.

### Response Format
```
Spend: {total}/mo  Allocated: {% tagged}
Top actions:
1. {action} | est. {savings}/mo | risk {low|med|high} | owner {team}
Commitment coverage target: {%} of steady baseline
Unit metric: {cost per X}
```

## FinOps Lifecycle

| Phase | Goal | Concrete outputs |
|-------|------|------------------|
| Inform | Everyone sees their cost | Mandatory tags (`team`, `service`, `env`, `cost-center`), cost dashboards, anomaly alerts |
| Optimize | Remove waste, lower rates | Right-sizing, schedules, spot, storage tiering, commitments |
| Operate | Keep it that way | Budgets per team, cost review in architecture/PR review, unit-cost KPIs |

## Optimization Levers (do in this order)

1. **Delete waste** — unattached volumes/disks, old snapshots, idle load balancers, unused public IPs, orphaned dev environments. Lowest risk, immediate.
2. **Schedule non-prod** — stop dev/staging outside working hours. A 12h×5d schedule runs 60 of 168 hours per week (~64% less runtime).
3. **Right-size** — use 2–4 weeks of utilization (CPU, memory, network; P95/P99 not averages). Tools: AWS Compute Optimizer, GCP Recommender, Azure Advisor. Prefer newer instance generations and ARM (Graviton/Axion/Cobalt-class) where software supports it.
4. **Elasticity** — autoscale on load; scale to zero for event-driven workloads.
5. **Spot/preemptible** — for stateless, checkpointed or retryable work (CI runners, batch, queue workers). Interruption notice: ~2 minutes on AWS, ~30 seconds on GCP and Azure. Diversify instance types and use capacity-optimized allocation.
6. **Commitments** — Savings Plans/RIs (AWS), Committed Use Discounts (GCP), Reservations/Savings Plans (Azure) for the steady baseline only, after right-sizing. Discounts depend on term, payment option and service; check the provider pricing page rather than quoting a fixed %.
7. **Storage tiering** — lifecycle policies to infrequent-access/archive classes; mind retrieval fees and minimum storage durations.

## Data Transfer (often the hidden line item)
- Internet egress is billed; ingress usually is not.
- Cross-AZ traffic is billed on AWS in both directions; chatty services and replicas across AZs add up.
- NAT gateway processing charges apply to traffic through it — use VPC gateway endpoints (S3, DynamoDB) or interface endpoints for high-volume AWS API traffic.
- Serve static assets via CDN; keep analytics/ETL in the same region as the data.
- Cross-region replication doubles storage and adds transfer — justify with RPO/RTO.

## Allocation & Accountability

```yaml
# Example tag policy (enforce via AWS Organizations tag policies / Azure Policy / GCP org policy + CI checks)
required_tags:
  team: "^[a-z-]+$"
  service: "^[a-z0-9-]+$"
  env: "^(prod|staging|dev|sandbox)$"
  cost-center: "^CC-[0-9]{4}$"
```
- Showback first (report), chargeback later (bill teams) once allocation >90% is trusted.
- Shared costs (networking, observability, platform) split by a documented key (usage, headcount or revenue).

## Unit Economics
Define one or two unit metrics tied to revenue drivers: cost per 1k requests, per active tenant, per order, per GB processed. Track trend per release; a rising unit cost with flat traffic is a regression.

## Anomaly Detection
- Enable provider anomaly detection (AWS Cost Anomaly Detection, GCP budget alerts, Azure cost alerts) per account/team.
- Budget alerts at 50/80/100% forecasted; route to the owning team channel, not finance only.

## Anti-Patterns
| Anti-pattern | Fix |
|--------------|-----|
| Buying 3-year commitments before right-sizing | Right-size, then commit to the stable floor |
| Optimizing averages | Size on P95/P99 plus headroom |
| Spot for stateful single instances | Spot only for interruptible work |
| Untagged resources | Block creation via policy-as-code |
| Finance-only cost ownership | Engineering teams own their spend dashboards |

## Checklist
- [ ] Tag coverage measured and enforced
- [ ] Non-prod schedules in place
- [ ] Right-sizing reviewed monthly
- [ ] Commitment coverage/utilization tracked
- [ ] Data-transfer paths reviewed (NAT, cross-AZ, egress)
- [ ] Unit cost metric on a dashboard

## References
- `skills/enterprise/cost-governance/SKILL.md` — budget governance and approvals
- `skills/devops/finops/SKILL.md` — FinOps tooling in delivery pipelines

### Shared References
Additional reference material for this skill, stored outside its own `references/` folder.

- `../references/finops_optimization.md` — FinOps: Cloud Financial Operations & Optimization

Wersje

W bazie nie ma innych skilli o nazwie enterprise-finops-cloud-cost.

Migawka bazy z 13.11.2026 · wersja paczki 8b5874f5db78 (skrót SHA-256 całej paczki)

Pobrania w ostatnich 30 dniach: 0 · ranking

Zgłoś problem z tym skillem: szkodliwa treść, błędny opis albo prośba autora o usunięcie.

Podobne skille

  • 100cloud-cost-finops-optimization
    Prowadzi przez ustalenie widoczności kosztów chmury i obniżanie wydatków w AWS, Azure i GCP: tagowanie, rightsizing, rabaty zobowiązaniowe (Reserved Instances, Savings Plans) oraz showback i chargeback FinOps.
  • 90finops-analyst
    Analizuje koszty chmury na AWS, GCP i Azure: przydział kosztów, rightsizing, rezerwacje, budżety, anomalie i koszt jednostkowy.
  • 100cloud-cost-optimization
    Znajduje i obniża wydatki na AWS, Azure, GCP i Kubernetes: dobór rozmiarów, nieużywane zasoby, rabaty i transfer danych.
  • 100aws-cost-optimizer
    Systematycznie obniża rachunek AWS: right-sizing przewymiarowanych zasobów, usuwanie bezczynnych i osieroconych, Savings Plans i Reserved Instances oraz tagowanie do alokacji kosztów.
  • 90enterprise-cost-governance
    Wdraża zarządzanie kosztami chmury: alokację, budżety, rozliczanie wewnętrzne, wykrywanie anomalii i tagowanie zasobów.
  • 90aws-cost-optimization-governor
    Przegląda koszty AWS i FinOps: Cost Explorer, Budgets, Compute Optimizer, Savings Plans, tagowanie, rightsizing, nieużywane zasoby, transfer danych i ryzyko prognozy.

Uwagi i zgłoszenia

Wiadomość trafi do Bartosza Sokolińskiego. Adres e-mail jest potrzebny tylko do odpowiedzi. Szczegóły w polityce prywatności.