optimizing-costs

Name: optimizing-costs
Author: ancoleman

48installs59forks

How to install optimizing-costs?

ancoleman/ai-design-components/optimizing-costs

$npx -y skills add ancoleman/ai-design-components --skill optimizing-costs

Installs into the current project.

Summary

Optimize cloud infrastructure costs through FinOps practices, commitment discounts, right-sizing, and automated cost management. Use when reducing cloud spend, implementing budget controls, or establishing cost visibility across AWS, Azure, GCP, and Kubernetes environments.

Files · 1

View on GitHub

SKILL.md

1	# Cost Optimization
2
3	## Purpose
4
5	Cloud cost optimization transforms uncontrolled spending into strategic resource allocation through the FinOps lifecycle: Inform, Optimize, and Operate. This skill provides decision frameworks for commitment-based discounts (Reserved Instances, Savings Plans), right-sizing strategies, Kubernetes cost management, and automated cost governance across multi-cloud environments.
6
7	## When to Use This Skill
8
9	Invoke cost-optimization when:
10	- Reducing cloud spend by 15-40% through systematic optimization
11	- Implementing cost visibility dashboards and allocation tracking
12	- Establishing budget alerts and anomaly detection
13	- Optimizing Kubernetes resource requests and cluster efficiency
14	- Managing Reserved Instances, Savings Plans, or Committed Use Discounts
15	- Automating idle resource cleanup and right-sizing recommendations
16	- Setting up showback/chargeback models for internal teams
17	- Preventing cost overruns through CI/CD cost estimation (Infracost)
18	- Responding to finance team requests for cloud cost reduction
19
20	## FinOps Principles
21
22	### The FinOps Lifecycle
23
24	```
25	┌─────────────────────────────────────────────────────┐
26	│ INFORM → OPTIMIZE → OPERATE (continuous loop) │
27	│ ↓ ↓ ↓ │
28	│ Visibility Action Automation │
29	└─────────────────────────────────────────────────────┘
30	```
31
32	Inform Phase: Establish cost visibility
33	- Enable cost allocation tags (Owner, Project, Environment)
34	- Deploy real-time cost dashboards for engineering teams
35	- Integrate cloud billing data (AWS CUR, Azure Consumption API, GCP BigQuery)
36	- Set up Kubernetes cost monitoring (Kubecost, OpenCost)
37
38	Optimize Phase: Take action on cost drivers
39	- Purchase commitment-based discounts (40-72% savings)
40	- Right-size over-provisioned resources (target 60-80% utilization)
41	- Implement spot/preemptible instances for fault-tolerant workloads
42	- Clean up idle resources (unattached volumes, old snapshots)
43
44	Operate Phase: Automate and govern
45	- Budget alerts with cascading notifications (50%, 75%, 90%, 100%)
46	- Automated cleanup scripts for idle resources
47	- CI/CD cost estimation to prevent surprise increases
48	- Continuous monitoring with anomaly detection
49
50	### Core FinOps Principles
51
52	1. Collaboration: Cross-functional teams (finance, engineering, operations, product)
53	2. Accountability: Teams own the cost of their services
54	3. Transparency: All costs visible and understandable to stakeholders
55	4. Optimization: Continuous improvement of cost efficiency
56
57	For detailed FinOps maturity models and organizational structures, see `references/finops-foundations.md`.
58
59	## Cost Optimization Strategies
60
61	### 1. Commitment-Based Discounts
62
63	Reserved Instances (RIs): 40-72% discount for 1-3 year commitments
64	- Standard RI: Instance type locked, highest discount (60% for 3-year)
65	- Convertible RI: Flexible instance types, moderate discount (54% for 3-year)
66	- Use for: Databases (RDS, ElastiCache), stable production EC2 workloads
67
68	Savings Plans: Flexible compute commitments
69	- Compute Savings Plans: Applies to EC2, Fargate, Lambda (54% discount for 3-year)
70	- EC2 Instance Savings Plans: Tied to instance family (66% discount for 3-year)
71	- Use for: Workloads that change instance types or regions
72
73	GCP Committed Use Discounts (CUDs): 25-70% discount
74	- Resource-based CUDs: Commit to vCPU, memory, GPUs
75	- Spend-based CUDs: Commit to dollar amount (flexible)
76	- Sustained Use Discounts: Automatic 20-30% discount for sustained usage (no commitment)
77
78	Decision Framework:
79	```
80	Reserve when:
81	├─ Workload is production-critical (24/7 uptime required)
82	├─ Usage is predictable (stable baseline over 6+ months)
83	├─ Architecture is stable (unlikely to change instance types)
84	└─ Financial commitment acceptable (1-3 year lock-in)
85
86	Use On-Demand when:
87	├─ Development/testing environments
88	├─ Unpredictable spiky workloads
89	├─ Short-term projects (<6 months)
90	└─ Evaluating new instance types
91	```
92
93	For detailed commitment strategies and RI coverage analysis, see `references/commitment-strategies.md`.
94
95	### 2. Spot and Preemptible Instances
96
97	Discount: 70-90% off on-demand pricing (interruptible with 2-minute warning)
98
99	Use Spot For: CI/CD workers, batch jobs, ML training (with checkpointing), Kubernetes workers, data analytics
100	Avoid Spot For: Stateful databases, real-time services, long-running jobs without checkpointing
101
102	Best Practices:
103	- Diversify instance types and spread across Availability Zones
104	- Implement graceful shutdown handlers
105	- Auto-fallback to on-demand when capacity unavailable
106	- Kubernetes: Mix 70% spot + 30% on-dema

People who liked this also viewed

6 picks

Type

optimizing-costs

48installs59forks

View on GitHub

How to install optimizing-costs?

ancoleman/ai-design-components/optimizing-costs

$npx -y skills add ancoleman/ai-design-components --skill optimizing-costs

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# Cost Optimization
2
3	## Purpose
4
5	Cloud cost optimization transforms uncontrolled spending into strategic resource allocation through the FinOps lifecycle: Inform, Optimize, and Operate. This skill provides decision frameworks for commitment-based discounts (Reserved Instances, Savings Plans), right-sizing strategies, Kubernetes cost management, and automated cost governance across multi-cloud environments.
6
7	## When to Use This Skill
8
9	Invoke cost-optimization when:
10	- Reducing cloud spend by 15-40% through systematic optimization
11	- Implementing cost visibility dashboards and allocation tracking
12	- Establishing budget alerts and anomaly detection
13	- Optimizing Kubernetes resource requests and cluster efficiency
14	- Managing Reserved Instances, Savings Plans, or Committed Use Discounts
15	- Automating idle resource cleanup and right-sizing recommendations
16	- Setting up showback/chargeback models for internal teams
17	- Preventing cost overruns through CI/CD cost estimation (Infracost)
18	- Responding to finance team requests for cloud cost reduction
19
20	## FinOps Principles
21
22	### The FinOps Lifecycle
23
24	```
25	┌─────────────────────────────────────────────────────┐
26	│ INFORM → OPTIMIZE → OPERATE (continuous loop) │
27	│ ↓ ↓ ↓ │
28	│ Visibility Action Automation │
29	└─────────────────────────────────────────────────────┘
30	```
31
32	Inform Phase: Establish cost visibility
33	- Enable cost allocation tags (Owner, Project, Environment)
34	- Deploy real-time cost dashboards for engineering teams
35	- Integrate cloud billing data (AWS CUR, Azure Consumption API, GCP BigQuery)
36	- Set up Kubernetes cost monitoring (Kubecost, OpenCost)
37
38	Optimize Phase: Take action on cost drivers
39	- Purchase commitment-based discounts (40-72% savings)
40	- Right-size over-provisioned resources (target 60-80% utilization)
41	- Implement spot/preemptible instances for fault-tolerant workloads
42	- Clean up idle resources (unattached volumes, old snapshots)
43
44	Operate Phase: Automate and govern
45	- Budget alerts with cascading notifications (50%, 75%, 90%, 100%)
46	- Automated cleanup scripts for idle resources
47	- CI/CD cost estimation to prevent surprise increases
48	- Continuous monitoring with anomaly detection
49
50	### Core FinOps Principles
51
52	1. Collaboration: Cross-functional teams (finance, engineering, operations, product)
53	2. Accountability: Teams own the cost of their services
54	3. Transparency: All costs visible and understandable to stakeholders
55	4. Optimization: Continuous improvement of cost efficiency
56
57	For detailed FinOps maturity models and organizational structures, see `references/finops-foundations.md`.
58
59	## Cost Optimization Strategies
60
61	### 1. Commitment-Based Discounts
62
63	Reserved Instances (RIs): 40-72% discount for 1-3 year commitments
64	- Standard RI: Instance type locked, highest discount (60% for 3-year)
65	- Convertible RI: Flexible instance types, moderate discount (54% for 3-year)
66	- Use for: Databases (RDS, ElastiCache), stable production EC2 workloads
67
68	Savings Plans: Flexible compute commitments
69	- Compute Savings Plans: Applies to EC2, Fargate, Lambda (54% discount for 3-year)
70	- EC2 Instance Savings Plans: Tied to instance family (66% discount for 3-year)
71	- Use for: Workloads that change instance types or regions
72
73	GCP Committed Use Discounts (CUDs): 25-70% discount
74	- Resource-based CUDs: Commit to vCPU, memory, GPUs
75	- Spend-based CUDs: Commit to dollar amount (flexible)
76	- Sustained Use Discounts: Automatic 20-30% discount for sustained usage (no commitment)
77
78	Decision Framework:
79	```
80	Reserve when:
81	├─ Workload is production-critical (24/7 uptime required)
82	├─ Usage is predictable (stable baseline over 6+ months)
83	├─ Architecture is stable (unlikely to change instance types)
84	└─ Financial commitment acceptable (1-3 year lock-in)
85
86	Use On-Demand when:
87	├─ Development/testing environments
88	├─ Unpredictable spiky workloads
89	├─ Short-term projects (<6 months)
90	└─ Evaluating new instance types
91	```
92
93	For detailed commitment strategies and RI coverage analysis, see `references/commitment-strategies.md`.
94
95	### 2. Spot and Preemptible Instances
96
97	Discount: 70-90% off on-demand pricing (interruptible with 2-minute warning)
98
99	Use Spot For: CI/CD workers, batch jobs, ML training (with checkpointing), Kubernetes workers, data analytics
100	Avoid Spot For: Stateful databases, real-time services, long-running jobs without checkpointing
101
102	Best Practices:
103	- Diversify instance types and spread across Availability Zones
104	- Implement graceful shutdown handlers
105	- Auto-fallback to on-demand when capacity unavailable
106	- Kubernetes: Mix 70% spot + 30% on-dema

People who liked this also viewed

6 picks

Type