Skip to content
Skills Library
Programming · Testing

benchmark

/benchmark

使用此技能测量性能基线、在 PR 之前/之后检测回归,并比较堆栈替代方案。

Licence
MIT
Last repository change
21.06.2026

Assessment and security

92/ 100Class: good
Format25 / 25
Frontmatter with name and description, valid name, description up to 1,024 characters, name matching the folder.
Discoverability17 / 25
Description explains when to use the skill, has a reasonable length and uses third person.
Content35 / 35
At least 150 words and at most 500 lines, headings, step lists, examples or code.
Hygiene15 / 15
No placeholders such as TODO, no Windows paths, with a licence.
Riskno signals detected in the scan
Files in this package
1
Scripts
0
HTML comments
0
URLs in SKILL.md
0
Deprecation

No signals detected. Repository changed on 21.06.2026, without references to retired models or APIs.

Licence

MIT, open: use and modify while retaining author attribution.

What it does and when to use it

使用此技能测量性能基线、在 PR 之前/之后检测回归,并比较堆栈替代方案。

Original author description

使用此技能测量性能基线、在 PR 之前/之后检测回归,并比较堆栈替代方案。

How it works

  1. 何时使用
  2. 工作原理
  3. 输出
  4. 集成
Original SKILL.md headings · 89 lines, 222 words
SKILL.md preview
---
name: benchmark
description: 使用此技能测量性能基线、在 PR 之前/之后检测回归,并比较堆栈替代方案。
origin: ECC
---

# 基准测试 — 性能基线和回归检测

## 何时使用

- 在 PR 之前和之后测量性能影响
- 为项目设置性能基线
- 当用户报告"感觉慢"时
- 发布之前 — 确保满足性能目标
- 将您的堆栈与替代方案进行比较

## 工作原理

### 模式 1:页面性能

通过浏览器 MCP 测量真实浏览器指标:

```
1. 导航到每个目标 URL
2. 测量核心 Web 指标:
   - LCP(最大内容绘制)— 目标 < 2.5s
   - CLS(累积布局偏移)— 目标 < 0.1
   - INP(交互到下次绘制)— 目标 < 200ms
   - FCP(首次内容绘制)— 目标 < 1.8s
   - TTFB(首字节时间)— 目标 < 800ms
3. 测量资源大小:
   - 总页面权重(目标 < 1MB)
   - JS 包大小(目标 < 200KB gzipped)
   - CSS 大小
   - 图像权重
   - 第三方脚本权重
4. 计算网络请求数
5. 检查渲染阻塞资源
```

### 模式 2:API 性能

对 API 端点进行基准测试:

```
1. 每个端点命中 100 次
2. 测量:p50、p95、p99 延迟
3. 跟踪:响应大小、状态代码
4. 负载下测试:10 个并发请求
5. 与 SLA 目标比较
```

### 模式 3:构建性能

测量开发反馈循环:

```
1. 冷构建时间
2. 热重载时间 (HMR)
3. 测试套件持续时间
4. TypeScript 检查时间
5. Lint 时间
6. Docker 构建时间
```

### 模式 4:之前/之后比较

在更改之前和之后运行以测量影响:

```
/benchmark baseline    # 保存当前指标
# ... 进行更改 ...
/benchmark compare     # 与基线比较
```

输出:
```
| 指标 | 之前 | 之后 | 增量 | 判定 |
|--------|--------|-------|-------|---------|
| LCP | 1.2s | 1.4s | +200ms | 警告:WARN |
| Bundle | 180KB | 175KB | -5KB | ✓ 更好 |
| Build | 12s | 14s | +2s | 警告:WARN |
```

## 输出

将基线存储在 `.ecc/benchmarks/` 中作为 JSON。Git 跟踪以便团队共享基线。

## 集成

- CI:在每个 PR 上运行 `/benchmark compare`
- 与 `/canary-watch` 配对进行部署后监控
- 与 `/browser-qa` 配对进行完整的发货前检查清单

Versions

RepositoryScoreLicence
This page · aaione/everything-claude-code-zh92MIT
workerstack/repo100MIT
YuDefine/nuxt-supabase-starter100MIT
hoangsonww/Claude-Code-Agent-Monitor100MIT
RainingDaemons/agents100MIT
wwylon/devkit100MIT
serejaris/personal-corp-os100MIT
pekral/cursor-rules100MIT
peabrane/peapods100MIT

Database snapshot: 13.11.2026 · package version 5247e4843997 (SHA-256 digest of the entire package)

Downloads in the last 30 days: 0 · ranking

Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.

Similar skills

  • 100dstudio-bench
    Measure Device's performance before and after a change - which engine, core or app benchmark covers what, how to run a fair comparison, when to add a benchmark, and how to report the numbers in a pull request. Use for any change that could affect CPU, memory, latency, frame time or load time.
  • 100tech-stack-research
    [Architecture] Use when a workflow step or the user asks for tech-stack research: alternatives, trade-offs and user-validated recommendations.
  • 100tech-stack-research
    [Architecture] Use when a workflow step or the user asks for tech-stack research: alternatives, trade-offs and user-validated recommendations.
  • 100benchmark-suite-creator
    Create benchmark suite creator operations. Auto-activating skill for Performance Testing. Triggers on: benchmark suite creator, benchmark suite creator Part of the Performance Testing skill category. Use when working with benchmark suite creator functionality. Trigger with…
  • 100run-benchmark
    Run the paper's replay benchmark on Evolving Infra, score it and compare it with the reference results. Use when asked to reproduce the paper's numbers, benchmark a change, or produce the performance table of a pull request. Not for finding why a run failed or is slower than the…
  • 100kelin-benchmark
    报性能数字前先取证:瓶颈是什么、有没有调优、是否可复现、端到端是否有意义。至少交替跑 5 次,报中位数和范围,结论只允许 faster/slower/no measurable difference/inconclusive。用户说“快了多少”“性能对比”“有没有变慢”“benchmark”时使用。

Feedback and reports

Your message goes to Bartosz Sokoliński. An email address is needed only for a reply. See the privacy policy.