Performance capacity
Plan, diagnose, and verify performance budgets, latency targets, load tests, capacity estimates, bottleneck analysis, caching strategy, query efficiency, queue throughput, and regression gates. Use when a feature may be slow, a system must scale, a performance regression is suspected, or release readiness depends on throughput, cost, memory, CPU, or response time.From its SKILL.md
npx -y skills add majiayu000/spellbook --skill performance-capacityAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
SKILL.md
2.0 KB, 343 tokens by cl100k_base, as published. Nobody here has run it
Performance Capacity
Purpose
Use this skill to make performance measurable before optimizing. It turns vague "make it faster" work into budgets, probes, bottleneck hypotheses, and regression gates.
Baseline First
Before changing code, capture:
- User-facing operation or background job under test.
- Current p50/p95/p99 latency or throughput.
- Data size and concurrency assumptions.
- Resource limits: CPU, memory, IO, network, database, queue.
- Existing cache behavior and invalidation rules.
- Cost or quota constraints.
If no baseline can be gathered, state the nearest measurable proxy and its limitations.
Budget Design
Define budgets by surface:
| Surface | Examples |
|---|---|
| UI | TTI, interaction latency, bundle size, render count |
| API | p95 latency, error rate, DB query count, payload size |
| Jobs | throughput, max lag, retry cost, idempotency |
| Data | query plan, index coverage, backfill duration |
| Infra | CPU/RSS, concurrency, autoscaling, cost per request |
Optimization Rules
- Optimize the measured bottleneck, not the most familiar code.
- Prefer algorithmic, query, batching, and cache correctness fixes before capacity-only fixes.
- Define cache invalidation and stale-data tolerance.
- Add a regression test, benchmark, or dashboard check for risky paths.
- Do not trade correctness, authorization, or tenant isolation for speed.
Output Shape
operation:
baseline:
target_budget:
bottleneck_hypothesis:
measurement_plan:
optimization_options:
capacity_estimate:
regression_gate:
verification_commands:
What ships with it: 1 file
253 B alongside SKILL.md
agents/
- openai.yaml253 B
Gives 0 of the 12 instructions most performance cost skills give in 343 tokens
Counted across 797 of the 1,117 authors here whose files we hold, read 2026-09-06
- Check for product marketing context firstin 46 of 797, across 20 files
- Measure before optimizingin 31 of 797, across 25 files
- Profile first to identify the actual bottleneckin 23 of 797, across 22 files
- Verify your robots.txt allows AI crawlersin 21 of 797, across 12 files
- Import directly and avoid barrel filesin 19 of 797, across 15 files
- Spawn all runs in the same turnin 18 of 797, across 11 files
- Write a draft of the skillin 17 of 797, across 10 files
- Understand the user's intentin 17 of 797, across 10 files
- Use React.cache for per-request deduplicationin 16 of 797, across 11 files
- Profile before optimizingin 16 of 797, across 14 files
- Include specific numbers with sourcesin 15 of 797, across 8 files
- Add lazy loading to below-fold imagesin 15 of 797, across 10 files
Said here and by no other author read
- Capture current metrics before changing code
- State the nearest measurable proxy if no baseline exists
- Define budgets by surface
- Prefer algorithmic and cache correctness fixes
- Define cache invalidation and stale-data tolerance
- Add a regression test or benchmark for risky paths
Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.