cloudthinker-ai Capacity Planning Review Skill
Perform capacity planning for {{ service_name }} over {{ planning_horizon }}.
Workflow
Step 1 — Current State Assessment
Gather current utilization metrics:
CURRENT UTILIZATION SNAPSHOT
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Traffic:
- Current RPS (avg): ___
- Current RPS (peak): ___
- Peak-to-average ratio: ___
- Daily/weekly traffic patterns: [describe]
Compute:
- Instance type/count: ___
- CPU utilization (avg/p95): ___% / ___%
- Memory utilization (avg/p95): ___% / ___%
- Pod/container count: ___
Data:
- Database size: ___ GB
- Database connections (avg/max): ___ / ___
- Cache hit ratio: ___%
- Storage growth rate: ___ GB/month
Network:
- Bandwidth utilization: ___ Mbps
- Request latency (p50/p95/p99): ___ms / ___ms / ___ms
Step 2 — Traffic Forecasting
TRAFFIC PROJECTION
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Growth assumption: {{ growth_rate | "historical trend" }}
| Timeframe | Projected RPS (avg) | Projected RPS (peak) | Confidence |
|-----------|--------------------|--------------------|------------|
| Current | [actual] | [actual] | - |
| +1 month | [projected] | [projected] | HIGH |
| +3 months | [projected] | [projected] | MEDIUM |
| +6 months | [projected] | [projected] | LOW |
| +12 months| [projected] | [projected] | LOW |
Known traffic events:
- [ ] Seasonal peaks (e.g., Black Friday, end-of-quarter)
- [ ] Planned launches or campaigns
- [ ] Market expansion or new customer onboarding
Step 3 — Resource Projections
RESOURCE PROJECTIONS
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
| Resource | Current | +3 Months | +6 Months | Limit | Headroom |
|----------|---------|-----------|-----------|-------|----------|
| Compute instances | X | Y | Z | Max | % |
| CPU cores | X | Y | Z | Max | % |
| Memory (GB) | X | Y | Z | Max | % |
| DB connections | X | Y | Z | Max | % |
| DB storage (GB) | X | Y | Z | Max | % |
| Cache memory | X | Y | Z | Max | % |
| Network bandwidth | X | Y | Z | Max | % |
Step 4 — Bottleneck Analysis
Identify the first constraint that will be hit:
BOTTLENECK RANKING
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
[ ] Database connections — pool exhaustion before compute ceiling
[ ] CPU — compute-bound workload hits capacity
[ ] Memory — OOM kills or swap before CPU ceiling
[ ] Storage — disk fills before other limits
[ ] Network — bandwidth or connection limits
[ ] Rate limits — external API or service limits
[ ] Queue depth — message backlog grows unbounded
[ ] Third-party dependency — external service becomes bottleneck
First bottleneck: [resource] at [projected date]
Mitigation: [scaling action needed]
Step 5 — Scaling Strategy
SCALING PLAN
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
SHORT-TERM (0-3 months):
[ ] Adjust autoscaling thresholds (current: X%, target: Y%)
[ ] Increase max instance count / replica count
[ ] Optimize hot paths (profiling-driven)
[ ] Cache tuning (increase hit ratio)
MEDIUM-TERM (3-6 months):
[ ] Vertical scaling (larger instance types)
[ ] Database read replicas or sharding
[ ] CDN optimization
[ ] Async processing for non-critical paths
LONG-TERM (6-12 months):
[ ] Architecture changes (decompose monolith, event-driven)
[ ] Multi-region deployment
[ ] Database migration (scale-out solution)
[ ] Cost optimization (reserved instances, spot fleet)
Step 6 — Cost Projection
| Resource | Current Monthly | +3 Months | +6 Months | Notes |
|---|---|---|---|---|
| Compute | $X | $Y | $Z | [scaling plan] |
| Database | $X | $Y | $Z | [growth + replicas] |
| Storage | $X | $Y | $Z | [growth rate] |
| Network | $X | $Y | $Z | [egress growth] |
| Total | $X | $Y | $Z |
Step 7 — Action Items
| Action | Priority | Timeline | Owner | Cost Impact |
|---|---|---|---|---|
| [action] | P1/P2/P3 | [date] | [name] | $X/month |
Counter-Rationalizations
| Shortcut | Counter | Why |
|---|---|---|
| "We can skip some steps for this case" | Adapt the workflow steps, don't skip them | Skipped steps are where incidents and oversights originate |
| "The user seems to already know what to do" | Complete all workflow phases with the user | The workflow catches blind spots that experience alone misses |
| "This is a minor case, full process is overkill" | Scale the process down, don't turn it off | Minor cases become major when unstructured; the process scales, not disappears |
| "I'll fill in the details later" | Complete each section before moving on | Deferred details are forgotten; real-time capture is more accurate |
| "The template output isn't necessary" | Always produce the structured output format | Structured output enables comparison, audit trails, and handoff to other teams |
Output Format
Produce a capacity planning report with:
- Current state utilization snapshot with key metrics
- Traffic forecast with growth projections and confidence levels
- Resource projections table with headroom analysis
- Bottleneck analysis identifying first constraint and timeline
- Scaling plan with short/medium/long-term actions
- Cost projection with monthly cost trajectory