MCP HubMCP Hub
返回技能列表

profiling-application-performance

jeremylongshore
更新于 Today
59 次查看
712
74
712
在 GitHub 上查看
ai

关于

This skill enables Claude to profile application performance by analyzing CPU usage, memory consumption, and execution time to identify bottlenecks. Use it when developers request performance analysis, optimization recommendations, or investigation of specific metrics like memory leaks. It leverages the application-profiler plugin to provide code-level optimization suggestions.

快速安装

Claude Code

推荐
插件命令推荐
/plugin add https://github.com/jeremylongshore/claude-code-plugins-plus
Git 克隆备选方式
git clone https://github.com/jeremylongshore/claude-code-plugins-plus.git ~/.claude/skills/profiling-application-performance

在 Claude Code 中复制并粘贴此命令以安装该技能

技能文档

Overview

This skill empowers Claude to analyze application performance, pinpoint bottlenecks, and recommend optimizations. By leveraging the application-profiler plugin, it provides insights into CPU usage, memory allocation, and execution time, enabling targeted improvements.

How It Works

  1. Identify Application Stack: Determines the application's technology (e.g., Node.js, Python, Java).
  2. Locate Entry Points: Identifies main application entry points and critical execution paths.
  3. Analyze Performance Metrics: Examines CPU usage, memory allocation, and execution time to detect bottlenecks.
  4. Generate Profile: Compiles the analysis into a comprehensive performance profile, highlighting areas for optimization.

When to Use This Skill

This skill activates when you need to:

  • Analyze application performance for bottlenecks.
  • Identify CPU-intensive operations and memory leaks.
  • Optimize application execution time.

Examples

Example 1: Identifying Memory Leaks

User request: "Analyze my Node.js application for memory leaks."

The skill will:

  1. Activate the application-profiler plugin.
  2. Analyze the application's memory allocation patterns.
  3. Generate a profile highlighting potential memory leaks.

Example 2: Optimizing CPU Usage

User request: "Profile my Python script and find the most CPU-intensive functions."

The skill will:

  1. Activate the application-profiler plugin.
  2. Analyze the script's CPU usage.
  3. Generate a profile identifying the functions consuming the most CPU time.

Best Practices

  • Code Instrumentation: Ensure the application code is instrumented for accurate profiling.
  • Realistic Workloads: Use realistic workloads during profiling to simulate real-world scenarios.
  • Iterative Optimization: Apply optimizations iteratively and re-profile to measure improvements.

Integration

This skill can be used in conjunction with code editing plugins to implement the recommended optimizations directly within the application's source code. It can also integrate with monitoring tools to track performance improvements over time.

GitHub 仓库

jeremylongshore/claude-code-plugins-plus
路径: backups/skills-batch-20251204-000554/plugins/performance/application-profiler/skills/application-profiler
aiautomationclaude-codedevopsmarketplacemcp

相关推荐技能

sglang

SGLang是一个专为LLM设计的高性能推理框架,特别适用于需要结构化输出的场景。它通过RadixAttention前缀缓存技术,在处理JSON、正则表达式、工具调用等具有重复前缀的复杂工作流时,能实现极速生成。如果你正在构建智能体或多轮对话系统,并追求远超vLLM的推理性能,SGLang是理想选择。

查看技能

evaluating-llms-harness

测试

该Skill通过60+个学术基准测试(如MMLU、GSM8K等)评估大语言模型质量,适用于模型对比、学术研究及训练进度追踪。它支持HuggingFace、vLLM和API接口,被EleutherAI等行业领先机构广泛采用。开发者可通过简单命令行快速对模型进行多任务批量评估。

查看技能

langchain

LangChain是一个用于构建LLM应用程序的框架,支持智能体、链和RAG应用开发。它提供多模型提供商支持、500+工具集成、记忆管理和向量检索等核心功能。开发者可用它快速构建聊天机器人、问答系统和自主代理,适用于从原型验证到生产部署的全流程。

查看技能

llamaguard

其他

LlamaGuard是Meta推出的7-8B参数内容审核模型,专门用于过滤LLM的输入和输出内容。它能检测六大安全风险类别(暴力/仇恨、性内容、武器、违禁品、自残、犯罪计划),准确率达94-95%。开发者可通过HuggingFace、vLLM或Sagemaker快速部署,并能与NeMo Guardrails集成实现自动化安全防护。

查看技能