cloudthinker-ai Spot.io / Ocean Management Skill
Manage and optimize cloud infrastructure costs with Spot.io and Ocean for Kubernetes.
MANDATORY: Discovery-First Pattern
Always list accounts and clusters before querying specific resources.
Phase 1: Discovery
#!/bin/bash
spot_api() {
local method="${1:-GET}"
local endpoint="$2"
local data="${3:-}"
local base_url="https://api.spotinst.io"
if [ -n "$data" ]; then
curl -s -X "$method" \
-H "Authorization: Bearer $SPOT_TOKEN" \
-H "Content-Type: application/json" \
"${base_url}/${endpoint}" -d "$data"
else
curl -s -X "$method" \
-H "Authorization: Bearer $SPOT_TOKEN" \
"${base_url}/${endpoint}"
fi
}
echo "=== Spot.io Account Info ==="
spot_api GET "setup/account" | jq -r '.response.items[] | "\(.accountId)\t\(.name)"' | column -t
echo ""
echo "=== Ocean Clusters ==="
spot_api GET "ocean/aws/k8s/cluster" | jq -r '
.response.items[] | "\(.id)\t\(.name)\t\(.controllerClusterId)\t\(.region)"
' | column -t
echo ""
echo "=== Elastigroup Summary ==="
spot_api GET "aws/ec2/group" | jq -r '
.response.items[] | "\(.id)\t\(.name)\t\(.capacity.target) instances\t\(.strategy.risk)% spot"
' | column -t | head -20
Core Helper Functions
#!/bin/bash
spot_api() {
local method="${1:-GET}"
local endpoint="$2"
local data="${3:-}"
if [ -n "$data" ]; then
curl -s -X "$method" -H "Authorization: Bearer $SPOT_TOKEN" \
-H "Content-Type: application/json" \
"https://api.spotinst.io/${endpoint}" -d "$data"
else
curl -s -X "$method" -H "Authorization: Bearer $SPOT_TOKEN" \
"https://api.spotinst.io/${endpoint}"
fi
}
spot_ocean() {
local endpoint="$1"
spot_api GET "ocean/aws/k8s/${endpoint}"
}
Output Rules
- TOKEN EFFICIENCY: Target ≤50 lines per output
- Use jq to filter API responses to relevant cost and scaling fields
- Round savings percentages and costs to 2 decimal places
Common Operations
Cluster Cost Optimization Report
#!/bin/bash
CLUSTER_ID="${1:?Ocean cluster ID required}"
echo "=== Ocean Cluster Cost Summary ==="
spot_api GET "ocean/aws/k8s/cluster/${CLUSTER_ID}/aggregatedCosts?startTime=$(date -u -d '7 days ago' +%Y-%m-%d)&endTime=$(date -u +%Y-%m-%d)" | jq '{
totalCost: .response.items[0].totalCost,
spotCost: .response.items[0].spotCost,
onDemandCost: .response.items[0].onDemandCost,
savings: .response.items[0].savings,
savingsPercentage: .response.items[0].savingsPercentage
}'
echo ""
echo "=== Cost by Namespace ==="
spot_api GET "ocean/aws/k8s/cluster/${CLUSTER_ID}/aggregatedCosts?startTime=$(date -u -d '7 days ago' +%Y-%m-%d)&endTime=$(date -u +%Y-%m-%d)&groupBy=namespace" | jq -r '
.response.items[] |
"\(.namespace)\t$\(.totalCost | . * 100 | round / 100)\tSavings: \(.savingsPercentage | . * 100 | round / 100)%"
' | sort -t'$' -k2 -rn | column -t | head -15
Workload Right-Sizing
#!/bin/bash
CLUSTER_ID="${1:?Ocean cluster ID required}"
echo "=== Right-Sizing Recommendations ==="
spot_api GET "ocean/aws/k8s/cluster/${CLUSTER_ID}/rightSizing/suggestion" | jq -r '
.response.items[] |
"\(.namespace)/\(.workloadName)\tCPU: \(.currentCpu)→\(.suggestedCpu)\tMem: \(.currentMemory)→\(.suggestedMemory)\tSavings: $\(.monthlySavings | . * 100 | round / 100)/mo"
' | sort -t'$' -k4 -rn | column -t | head -20
Savings Analysis
#!/bin/bash
echo "=== Overall Savings Summary ==="
spot_api GET "ocean/aws/k8s/cluster" | jq -r '.response.items[].id' | while read cid; do
NAME=$(spot_api GET "ocean/aws/k8s/cluster/${cid}" | jq -r '.response.items[0].name')
SAVINGS=$(spot_api GET "ocean/aws/k8s/cluster/${cid}/aggregatedCosts?startTime=$(date -u -d '30 days ago' +%Y-%m-%d)&endTime=$(date -u +%Y-%m-%d)" | jq '.response.items[0]')
echo "$NAME ($cid)"
echo "$SAVINGS" | jq '"\tTotal: $\(.totalCost | . * 100 | round / 100)\tSaved: $\(.savings | . * 100 | round / 100)\t(\(.savingsPercentage | . * 100 | round / 100)%)"'
done
echo ""
echo "=== Spot Instance Usage ==="
spot_api GET "aws/ec2/group" | jq -r '
.response.items[] |
"\(.name)\tTarget: \(.capacity.target)\tSpot: \(.strategy.risk)%\tFallback: \(.strategy.fallbackToOd // false)"
' | column -t | head -15
Node Scaling Events
#!/bin/bash
CLUSTER_ID="${1:?Ocean cluster ID required}"
echo "=== Recent Scaling Events ==="
spot_api GET "ocean/aws/k8s/cluster/${CLUSTER_ID}/log?fromDate=$(date -u -d '24 hours ago' +%Y-%m-%dT%H:%M:%SZ)&limit=30" | jq -r '
.response.items[] |
"\(.createdAt[0:16])\t\(.message[0:80])"
' | head -20
echo ""
echo "=== Current Node Pool Status ==="
spot_api GET "ocean/aws/k8s/cluster/${CLUSTER_ID}/nodes" | jq -r '
.response.items[] |
"\(.instanceId)\t\(.instanceType)\t\(.lifeCycle)\t\(.availabilityZone)\t\(.nodeStatus)"
' | column -t | head -20
Elastigroup Management
#!/bin/bash
GROUP_ID="${1:?Elastigroup ID required}"
echo "=== Elastigroup Status ==="
spot_api GET "aws/ec2/group/${GROUP_ID}" | jq '{
name: .response.items[0].name,
target: .response.items[0].capacity.target,
min: .response.items[0].capacity.minimum,
max: .response.items[0].capacity.maximum,
spotPercentage: .response.items[0].strategy.risk,
region: .response.items[0].region
}'
echo ""
echo "=== Active Instances ==="
spot_api GET "aws/ec2/group/${GROUP_ID}/status" | jq -r '
.response.items[] |
"\(.instanceId)\t\(.instanceType)\t\(.product)\t\(.availabilityZone)\t\(.privateIp)"
' | column -t | head -15
Safety Rules
- Read-only queries first: Always review current state before scaling or modifying groups
- Fallback to on-demand: Ensure
fallbackToOdis enabled before relying on spot instances - Capacity minimums: Never set minimum capacity to 0 for production workloads
- Gradual changes: When adjusting spot percentage, change incrementally (10-20% at a time)
Output Format
Present results as a structured report:
Managing Spot Io Report
═══════════════════════
Resources discovered: [count]
Resource Status Key Metric Issues
──────────────────────────────────────────────
[name] [ok/warn] [value] [findings]
Summary: [total] resources | [ok] healthy | [warn] warnings | [crit] critical
Action Items: [list of prioritized findings]
Target ≤50 lines of output. Use tables for multi-resource comparisons.
Anti-Hallucination Rules
- NEVER assume resource names — always discover via CLI/API in Phase 1 before referencing in Phase 2.
- NEVER fabricate metric names or dimensions — verify against the service documentation or
--helpoutput. - NEVER mix CLI commands between service versions — confirm which version/API you are targeting.
- ALWAYS use the discovery → verify → analyze chain — every resource referenced must have been discovered first.
- ALWAYS handle empty results gracefully — an empty response is valid data, not an error to retry.
Counter-Rationalizations
| Shortcut | Counter | Why |
|---|---|---|
| "I'll skip discovery and check known resources" | Always run Phase 1 discovery first | Resource names change, new resources appear — assumed names cause errors |
| "The user only asked for a quick check" | Follow the full discovery → analysis flow | Quick checks miss critical issues; structured analysis catches silent failures |
| "Default configuration is probably fine" | Audit configuration explicitly | Defaults often leave logging, security, and optimization features disabled |
| "Metrics aren't needed for this" | Always check relevant metrics when available | API/CLI responses show current state; metrics reveal trends and intermittent issues |
| "I don't have access to that" | Try the command and report the actual error | Assumed permission failures prevent useful investigation; actual errors are informative |
Common Pitfalls
- Token scope: Spot API tokens are account-scoped -- ensure correct account context
- Ocean vs Elastigroup: Ocean is Kubernetes-native; Elastigroup is for generic EC2 -- do not confuse APIs
- Spot interruptions: High spot percentage without on-demand fallback risks availability
- Right-sizing lag: Recommendations need 3-7 days of data to stabilize
- Region mismatch: Ensure cluster region matches when querying aggregated costs
- Controller connectivity: Ocean requires the Spot controller running in-cluster -- check controller pod health