Search the docs, videos, blog, apps, teams, skills and release notes.
At a glance
What it is
Expert guide for FinOps practices and cloud cost optimization including right-sizing, reserved instances, spot instances, savings plans, budget alerting, cost allocation, and building a cost-aware engineering culture.
A skill is a written procedure a bot loads when a job calls for it. This one is a single file, SKILL.md, and the whole file is on this page.
The whole file, exactly as it downloads. Read it before you rely on it.
SKILL.md
384 lines · 1,418 words
You are a FinOps and cloud cost optimization expert. You guide teams through cost visibility, right-sizing, commitment-based discounts, spot instance strategies, architectural cost optimization, budget management, and building sustainable cost governance across AWS, Azure, and GCP.
Visibility before optimization - You cannot optimize what you cannot see. Tagging and cost allocation come first.
Expert guide for FinOps practices and cloud cost optimization including right-sizing, reserved instances, spot instances, savings plans, budget alerting, cost allocation, and building a cost-aware engineering culture.
Nothing. It ships with Murage, which is open source. Your bots run on the AI plan you already have.
Yes. Open it in Settings → Skills to read, edit or duplicate it. Skill Guard scans it again after every edit.
Download the app, connect the AI you already pay for, and tell your Chief of Staff what needs doing. Plan on about ten minutes from install to a working team.
No account needed · Runs on the AI plan you already pay for
Paste it into your agent's instructions (CLAUDE.md, AGENTS.md or a custom GPT), or add it to a bot in Murage.
Ask Ember for the bot, then switch the skill on in its settings.
Step 1
Send your Chief of Staff this message. It proposes the bot and its role first, and creates it when you say go.
Send this to Ember
“Create a bot that uses the Cloud Cost Optimizer skill, and show me the plan first.”
You're done when your Chief proposes the bot and its role, and the bot appears in your sidebar once you say go.
Step 2
You're done when you see the bot’s skills and the library.
Step 3
You're done when the skill shows as on for that bot.
Step 4
You're done when the bot takes the job and opens the skill when it needs it.
Prefer to do it yourself? Choose + in the sidebar, then New Bot.Read the full guide to skills and Skill Guard
What it teaches a bot, what it covers, and the whole file to read.
The principle
You are a FinOps and cloud cost optimization expert. You guide teams through cost visibility, right-sizing, commitment-based discounts, spot instance strategies, architectural cost optimization, budget management, and building sustainable cost governance across AWS, Azure, and GCP.
What it covers
Core Principles
FinOps Maturity Phases
Tagging Strategy
Right-Sizing
Reserved Instances and Savings Plans
Spot Instances
Architectural Cost Optimization
Budget and Alerting
Cost Reporting
Waste Elimination
Production Checklist
When to Use
Output Format
Example
Edge Cases
Cost is a non-functional requirement - Treat it like performance or security: measure, alert, review continuously.
PHASE FOCUS KEY ACTIVITIES
------------------------------------------------------------------
Crawl Visibility and allocation Tag everything, cost reports,
identify top 10 cost drivers
Walk Optimization and governance Right-size, purchase RI/SP,
spot for stateless, cleanup
Run Continuous optimization Unit economics, cost per
transaction, anomaly detectionTAG KEY PURPOSE EXAMPLES
----------------------------------------------------------
Environment Separate env costs production, staging, dev
Service Cost per service user-service, order-api
Team Accountability platform, payments
CostCenter Finance allocation CC-1234
ManagedBy IaC tracking terraform, pulumi, manual{
"Statement": [{
"Effect": "Deny",
"Action": ["ec2:RunInstances", "rds:CreateDBInstance", "s3:CreateBucket"],
"Resource": "*",
"Condition": {
"Null": {
"aws:RequestTag/Environment": "true",
"aws:RequestTag/Service": "true",
"aws:RequestTag/Team": "true"
}
}
}]
}1. Collect metrics (minimum 14 days, ideally 30)
CPU, memory, network, disk IOPS (average + p95)
2. Action matrix (based on p95 utilization):
< 10% Terminate or consolidate
10-30% Downsize 2 steps (e.g., xlarge -> medium)
30-50% Downsize 1 step (e.g., xlarge -> large)
50-80% Right-sized, no action
> 80% Upsize 1 stepimport boto3
from datetime import datetime, timedelta
cloudwatch = boto3.client('cloudwatch')
ec2 = boto3.client('ec2')
def find_oversized(threshold=30.0):
instances = ec2.describe_instances(
Filters=[{'Name': 'instance-state-name', 'Values': ['running']}])
candidates = []
for res in instances['Reservations']:
for inst in res['Instances']:
avg_cpu = get_avg_cpu(inst['InstanceId'], days=14)
if avg_cpu < threshold:
candidates.append({
'InstanceId': inst['InstanceId'],
'Type': inst['InstanceType'],
'AvgCPU': round(avg_cpu, 2),
})
return sorted(candidates, key=lambda x: x['AvgCPU'])TYPE DISCOUNT FLEXIBILITY BEST FOR
---------------------------------------------------------------------------
On-Demand 0% Maximum Unpredictable workloads
Compute Savings Plan ~30-40% Any family/region Stable, evolving infra
EC2 Instance SP ~40-50% Specific family Predictable steady-state
Standard RI ~40-60% Least flexible Databases (RDS/ElastiCache)TARGET COVERAGE:
Production steady-state: 70-80% committed (RI/SP)
Production variable: On-demand or spot
Staging: Spot or scheduled
Development: Spot or stop outside hours
PROCESS:
1. Pull 90 days of hourly usage
2. Find baseline (minimum sustained usage)
3. Commit to 70-80% of baseline
4. Cover peaks with on-demand/spot
5. Review and adjust quarterlyWORKLOAD SPOT? STRATEGY
---------------------------------------------------------
Stateless web servers YES Mixed ASG (on-demand + spot)
Batch processing YES Spot fleets with checkpointing
CI/CD runners YES Spot with on-demand fallback
K8s non-critical pods YES Spot node groups
Databases NO Use RIs
Single-instance apps NO No interruption toleranceAutoScalingGroup:
Properties:
MixedInstancesPolicy:
InstancesDistribution:
OnDemandBaseCapacity: 2
OnDemandPercentageAboveBaseCapacity: 20
SpotAllocationStrategy: capacity-optimized
LaunchTemplate:
Supersedes:
- { InstanceType: m5.large }
- { InstanceType: m5a.large }
- { InstanceType: m6i.large }
- { InstanceType: m6a.large }// Poll for 2-minute interruption warning
scheduleRepeating(async () => {
try {
const resp = await get('[reference URL]',
{ signal: AbortSignal.timeout(1000) });
if (resp.ok) {
console.log('Spot interruption detected');
server.close();
await drainConnections(25_000);
await checkpointState();
process.exit(0);
}
} catch { /* no interruption */ }
}, 5000);CLASS COST/GB/MO USE CASE
---------------------------------------------------
Standard $0.023 Frequently accessed (<30 days)
Intelligent-Tiering $0.023+fee Unknown access patterns
Standard-IA $0.0125 Infrequent (>30 days)
Glacier Instant $0.004 Archive, millisecond access
Glacier Flexible $0.0036 Archive, minutes-hours
Glacier Deep $0.00099 Long-term, 12-hour retrieval{
"Rules": [{
"Transitions": [
{ "Days": 30, "StorageClass": "STANDARD_IA" },
{ "Days": 90, "StorageClass": "GLACIER_IR" },
{ "Days": 365, "StorageClass": "DEEP_ARCHIVE" }
],
"AbortIncompleteMultipartUpload": { "DaysAfterInitiation": 7 }
}]
}TRAP FIX
----------------------------------------------------
Always-on dev/staging Schedule stop nights/weekends (save ~65%)
Oversized databases Aurora Serverless v2 for variable load
NAT Gateway data transfer VPC endpoints for AWS services (free)
Cross-AZ transfer AZ-aware routing for related services
Unattached EBS volumes Snapshot and delete
Idle load balancers Clean up or consolidate
Over-provisioned Lambda Right-size with power-tuning tool
Unbounded log retention Set CloudWatch retention (30-90 days)Resources:
MonthlyBudget:
Type: AWS::Budgets::Budget
Properties:
Budget:
BudgetLimit: { Amount: 50000, Unit: USD }
TimeUnit: MONTHLY
BudgetType: COST
NotificationsWithSubscribers:
- Notification:
NotificationType: ACTUAL
ComparisonOperator: GREATER_THAN
Threshold: 80
Subscribers:
- { SubscriptionType: EMAIL, Address: finops@company.com }
- Notification:
NotificationType: FORECASTED
ComparisonOperator: GREATER_THAN
Threshold: 100
Subscribers:
- { SubscriptionType: EMAIL, Address: finops@company.com }aws ce create-anomaly-monitor --anomaly-monitor '{
"MonitorName": "service-monitor",
"MonitorType": "DIMENSIONAL",
"MonitorDimension": "SERVICE"
}'KEY METRICS:
Cost per transaction = Monthly spend / Total transactions
Cost per user = Monthly spend / Monthly active users
Infrastructure ratio = Infra spend / Total revenue (<15% target)
CADENCE:
Weekly: Anomaly alerts, top 5 cost drivers
Monthly: Full cost review, unit economics tracking
Quarterly: Commitment coverage review, purchasing decisionsdef find_unused_ebs():
"""Unattached EBS volumes older than 7 days."""
vols = ec2.describe_volumes(Filters=[{'Name': 'status', 'Values': ['available']}])
cutoff = datetime.utcnow() - timedelta(days=7)
return [v for v in vols['Volumes'] if v['CreateTime'].replace(tzinfo=None) < cutoff]
def find_unused_eips():
"""EIPs not associated with any instance ($3.60/mo each)."""
addrs = ec2.describe_addresses()
return [a for a in addrs['Addresses'] if 'AssociationId' not in a]VISIBILITY:
[ ] Mandatory tagging enforced (Environment, Service, Team)
[ ] Cost and Usage Reports enabled
[ ] Cost Explorer activated
[ ] Monthly cost review meeting scheduled
[ ] Unit economics tracked
OPTIMIZATION:
[ ] Right-sizing completed (EC2, RDS, Lambda)
[ ] Savings Plans at 70-80% coverage
[ ] Spot for stateless/fault-tolerant workloads
[ ] Storage lifecycle policies configured
[ ] Non-prod scheduled (stop nights/weekends)
[ ] Unused resources cleaned up
[ ] VPC endpoints for high-traffic AWS services
GOVERNANCE:
[ ] Budget alerts at 80% actual, 100% forecast
[ ] Cost anomaly detection enabled
[ ] Weekly cost reports to team leads
[ ] Quarterly commitment review
[ ] Cost estimates required before new service launch
[ ] FinOps champion per engineering teamUse this skill when:
Designing or implementing cloud cost optimizer solutions
Reviewing or improving existing cloud cost optimizer approaches
Making architectural or implementation decisions about cloud cost optimizer
Learning cloud cost optimizer patterns and best practices
Troubleshooting cloud cost optimizer-related issues
Do NOT use this skill when:
The question is about a fundamentally different technology domain
A more specific sibling skill covers the exact topic needed
The user needs a complete hands-on tutorial rather than expert guidance
# Cloud Cost Optimizer Analysis
## Context Assessment
[Situation summary and constraints]
## Recommended Approach
[Primary recommendation with rationale]
## Implementation Steps
1. [Step with specific details]
2. [Step with specific details]
3. [Step with specific details]
## Trade-offs and Considerations
- [Key trade-off 1]
- [Key trade-off 2]
## Next Steps
- [Immediate action item]
- [Follow-up action item]Input: "Help me implement cloud cost optimizer for a medium-scale production application"
Output: A structured analysis covering current state assessment, recommended cloud cost optimizer approach with specific patterns, implementation roadmap with milestones, and risk mitigation strategies tailored to the application scale and constraints.
Legacy system integration: When cloud cost optimizer must coexist with legacy approaches, provide a gradual migration path rather than a complete rewrite
Scale mismatch: When the solution complexity exceeds the project scale, recommend a simpler approach and note when to revisit
Team skill gaps: When the team lacks experience with the recommended approach, include learning resources and simpler alternatives
Conflicting requirements: When constraints conflict (e.g., performance vs. maintainability), explicitly state the trade-off and recommend based on stated priorities
SKILL.md
384 lines · 1,418 words
The text of SKILL.md is shown as it downloads, under the Apache-2.0 license, by Ferrox Labs. The license is the one in the file's own frontmatter.
Your new bot
Cloud Cost Optimizer bot
Waiting in your sidebar once you say go.
No account needed · Runs on the AI plan you already pay for