Token导航 LogoToken导航

Agent Skill · 研究检索

flutter-performance(Flutter 性能)

flutter-performance 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

flutter-performance 是研究检索类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为只读;许可证信息还不明确,第三方扫描已通过,按安装量在研究检索类的 73,675 个 Skill 里排第 4,842。

总安装 29,376周安装 1,189GitHub Stars 1,326下载 9,408核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
unknown

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:flutter-performance(Flutter 性能)
来源仓库:https://github.com/flutter/skills
仓库路径:skills/flutter-performance
安装命令:
npx skills add https://github.com/flutter/skills --skill flutter-performance
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/flutter/skills --skill flutter-performance

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题,通常会让助手先检索再总结。

通过系统分析和有针对性的优化,识别并消除 Flutter 应用程序中的性能瓶颈。

提供决策树来诊断 UI 线程、光栅 (GPU) 线程或两者上的卡顿,并针对每个线程提供特定修复 包括使用traceAction的集成测试模板 和时间线总结 建立绩效基线并衡量框架预算 涵盖 UI 优化(本地化状态、const 构造函数、StringBuffer 使用)和光栅优化(替换不透明度小部件,避免昂贵的裁剪) 解决了布局效率低下、使用 LayoutBuilder/OverlayEntry 进行框架破坏性更改以及通过 Chrome DevTools 进行特定于 Web 的分析的问题 强制约束:--profile 中的配置文件 仅模式,避免小部件上的运算符覆盖,并对长列表使用惰性构建器

来源仓库内路径:skills/flutter-performance

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 Codex、Claude、Cursor、Gemini CLI;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 npx skills add https://github.com/flutter/skills --skill flutter-performance;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它标注为只读,先用只读任务试一次,确认行为符合预期再扩大使用范围。

常见问题

6 条

flutter-performance 怎么安装?

在终端执行 npx skills add https://github.com/flutter/skills --skill flutter-performance。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

flutter-performance 支持哪些 AI 工具?

公开资料标注支持 Codex、Claude、Cursor、Gemini CLI。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

flutter-performance 需要什么权限?

权限级别是只读。该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

flutter-performance 是免费的吗?

公开资料没有标注明确许可证,收费和商用限制都无法确认。正式使用前请到来源仓库核对授权说明。

flutter-performance 用的人多吗?

公开数据是累计安装 29,376 次,近一周安装 1,189 次,下载 9,408 次,GitHub 1,326 星。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

flutter-performance 安全吗?

第三方扫描结果为通过,但这只覆盖已知的静态检查项,不能替代自己读一遍源码和安装命令。

值得选的地方和要先想清楚的地方

值得选的地方

  • 同时适配 4 种工具,换客户端不用重新找替代品。
  • 第三方扫描已通过,常见的高风险行为没有命中。
  • 来源仓库有 1,326 星,维护活跃度相对可查。
  • 标注为只读权限,试用阶段不容易误改本地文件。

先想清楚的地方

  • 许可证没有明确标注,能不能商用、能不能再分发都不确定。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 Codex,想直接扩展它能力的人。
  • 需要处理研究检索类任务、又不想每次都手写提示词的人。
  • 只想先试试效果、不希望它改动本地文件的人。

不太适合

  • 需要明确授权才能用到公司项目里的场景。
  • 需要厂商级技术支持和 SLA 的生产系统。

研究检索类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
find-skills

寻找技能 · Codex、Claude

31,199,81316,486操作浏览器MIT有提醒
self-improving-agent

自我改进 Agent · OpenClaw

16,464,0023,396接触敏感数据MIT-0通过
soultrace

灵魂追踪 · Codex、Claude

7,588,94412需要联网未明确有提醒
self-improving

自我完善 · OpenClaw

6,829,2841,046需要联网MIT-0通过
microsoft-foundry

微软代工厂 · Codex、Claude

6,614,144764调用外部服务未明确有提醒
flutter-performance

Flutter 性能 · 本页

29,3761,326只读未明确通过

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

Flutter Performance Optimization

Goal

Analyzes and optimizes Flutter application performance by identifying jank, excessive rebuilds, and expensive rendering operations. Implements best practices for UI rendering, state management, and layout constraints. Utilizes Flutter DevTools, Chrome DevTools (for web), and integration tests to generate actionable performance metrics, ensuring frames render within the strict 16ms budget.

Decision Logic

Evaluate the target application using the following decision tree to determine the optimization path:

  1. Is the goal to establish a performance baseline?

- Yes: Implement an integration test using traceAction and TimelineSummary. - No: Proceed to step 2.

  1. Is the application running on Web?

- Yes: Enable debugProfileBuildsEnabled and use Chrome DevTools Performance panel. - No: Run the app on a physical device in --profile mode and launch Flutter DevTools.

  1. Which thread is showing jank (red bars > 16ms) in the DevTools Performance View?

- UI Thread: Optimize build() methods, localize setState(), use const constructors, and replace string concatenation with StringBuffer. - Raster (GPU) Thread: Minimize saveLayer(), Opacity, Clip, and ImageFilter usage. Pre-cache complex images using RepaintBoundary. - Both: Start by optimizing the UI thread (Dart VM), as expensive Dart code often cascades into expensive rendering.

Instructions

  1. Establish a Performance Baseline To measure performance programmatically, create an integration test that records a performance timeline. STOP AND ASK THE USER: "Do you want to run a baseline integration test to capture timeline metrics before optimizing?" If yes, implement the following exact driver and test implementation: *test_driver/perf_driver.dart* (Immutable operation): import 'package:flutter_driver/flutter_driver.dart' as driver; import 'package:integration_test/integration_test_driver.dart'; Future<void> main() {return integrationDriver(responseDataCallback: (data) async {if (data!= null) {final timeline = driver.Timeline.fromJson(data['scrolling_timeline'] as Map<String, dynamic>,); final summary = driver.TimelineSummary.summarize(timeline); await summary.writeTimelineToFile('scrolling_timeline', pretty: true, includeSummary: true,);}},);} *integration_test/scrolling_test.dart*: import 'package:flutter/material.dart'; import 'package:flutter_test/flutter_test.dart'; import 'package:integration_test/integration_test.dart'; import 'package:your_package/main.dart'; void main() {final binding = IntegrationTestWidgetsFlutterBinding.ensureInitialized(); testWidgets('Performance profiling test', (tester) async {await tester.pumpWidget(const MyApp()); final listFinder = find.byType(Scrollable); final itemFinder = find.byKey(const ValueKey('target_item')); await binding.traceAction(() async {await tester.scrollUntilVisible(itemFinder, 500.0, scrollable: listFinder,);}, reportKey: 'scrolling_timeline');});} Run the test using: flutter drive --driver=test_driver/perf_driver.dart --target=integration_test/scrolling_test.dart --profile --no-dds
  2. Optimize UI Thread (Build Costs) If the UI thread exceeds 8ms per frame, refactor the widget tree: // BAD: Rebuilds entire widget tree setState(() {_counter++;}); // GOOD: Encapsulated state class CounterWidget extends StatefulWidget {...} // Inside CounterWidget: setState(() {_counter++;}); // GOOD: Efficient string building final buffer = StringBuffer(); for (int i = 0; i < 1000; i++) {buffer.write('Item $i');} final result = buffer.toString();

- Localize State: Move setState calls as low in the widget tree as possible. - Use const: Apply const constructors to short-circuit rebuild traversals. - String Building: Replace + operators in loops with StringBuffer.

  1. Optimize Raster Thread (Rendering Costs) If the Raster thread exceeds 8ms per frame, eliminate expensive painting operations: // BAD: Expensive Opacity widget Opacity(opacity: 0.5, child: Container(color: Colors.red),) // GOOD: Semitransparent color Container(color: Colors.red.withOpacity(0.5))

- Replace Opacity widgets with semitransparent colors where possible. - Replace Opacity in animations with AnimatedOpacity or FadeInImage. - Avoid Clip.antiAliasWithSaveLayer. Use borderRadius properties on containers instead of explicit clipping widgets.

  1. Fix Layout and Intrinsic Passes Identify and remove excessive layout passes caused by intrinsic operations (e.g., asking all children for their size before laying them out).

- Use lazy builders (ListView.builder, GridView.builder) for long lists. - Avoid ShrinkWrap: true on scrollables unless absolutely necessary.

  1. Handle Framework Breaking Changes (Validate-and-Fix) Ensure the application complies with recent Flutter optimization changes regarding LayoutBuilder and OverlayEntry. These widgets no longer rebuild implicitly. // FIX: Explicit setState for Overlay/Route changes final newLabel = await Navigator.pushNamed(context, '/bar'); setState(() {buttonLabel = newLabel;});

- Validate: Check if LayoutBuilder or OverlayEntry UI fails to update. - Fix: Wrap the state modification triggering the update in an explicit setState.

  1. Web-Specific Profiling If profiling for Web, inject timeline events into Chrome DevTools by adding these flags to main() before runApp(): import 'package:flutter/widgets.dart'; import 'package:flutter/rendering.dart'; void main() {debugProfileBuildsEnabled = true; debugProfileBuildsEnabledUserWidgets = true; debugProfileLayoutsEnabled = true; debugProfilePaintsEnabled = true; runApp(const MyApp());} STOP AND ASK THE USER: "Have you captured the Chrome DevTools performance profile? Please share the timeline event bottlenecks if you need specific refactoring."

Constraints

  • NEVER profile performance in Debug mode. Always use --profile mode on a physical device.
  • NEVER override operator == on Widget objects. It results in O(N²) behavior and degrades performance. Rely on const caching instead.
  • NEVER put a subtree in an AnimatedBuilder that does not depend on the animation. Build the static part once and pass it as the child parameter.
  • DO NOT use constructors with a concrete List of children (e.g., Column, ListView) if most children are off-screen. Always use .builder constructors for lazy loading.
  • DO NOT use saveLayer() unless absolutely necessary (e.g., dynamic overlapping shapes with transparency). Precalculate and cache static overlapping shapes.

平台分布

Codex

35.09%
按下载量换算3,301

Claude

34.74%
按下载量换算3,268

Cursor

18.39%
按下载量换算1,730

Gemini CLI

9.13%
按下载量换算859

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

只读

该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills