本提示词提供一套系统化源码分析流程，从项目概览、架构理解、核心流程深入、细节深入到总结应用五个阶段，帮助深入理解开源项目的架构与核心逻辑，掌握设计思想并迁移到实际开发。适用于源码学习、技术调研与架构设计等场景。

中文版提示词

目标：深入理解一个开源项目的架构与核心逻辑，掌握其设计思想并能应用到实际开发中。适用对象：开发者、技术研究者、架构师、学生。使用方式：按步骤执行，每步可结合代码阅读、调试、绘图工具辅助。

一、项目概览：查看README/文档了解背景、功能定位、使用场景；确认版本与分支，明确分析的是哪个版本；检查运行环境要求（依赖库、编译方式、启动命令）；通过示例程序或API调用快速体验功能。提示：先建立整体认知，不要一开始就钻进代码。

二、架构理解，从四个维度入手：1. 目录结构分析（列出主要目录、分析模块划分逻辑、标注关键文件）；2. 技术栈识别（语言、框架与库、架构模式）；3. 架构图绘制（用draw.io/PlantUML绘制模块关系图，明确组件间调用关系、数据流向、通信机制）；4. 核心流程梳理（定位主线业务流程、识别关键阶段、绘制流程图）。

三、核心流程深入：定位入口文件与初始化过程；追踪核心函数调用链，用调试器或日志辅助跟踪执行路径，关注异常处理、资源释放、锁机制；分析数据流（数据如何传递、标记关键数据结构、理解数据生命周期）；组件分析（各模块职责、接口定义、是否存在抽象层）。

四、细节深入：对复杂函数或模块逐行解读；分析性能优化点（内存池、缓存、异步处理）；探究边界条件处理（空指针、超时、并发冲突）；查看错误码与日志输出机制。

五、总结应用：总结设计亮点（可扩展性、可维护性、高性能设计、设计模式应用）；发现潜在问题（内存泄漏风险、竞态条件、配置缺失）；提出改进建议（如何优化逻辑、如何支持新功能）；迁移应用（将设计思路迁移到自己的项目，撰写技术博客或分享文档）。

英文版提示词

Goal: deeply understand the architecture and core logic of an open-source project, grasp its design philosophy, and apply it in real development. Audience: developers, technical researchers, architects, students. Usage: follow the steps in order, using code reading, debugging, and diagramming tools as aids.

1. Project overview: check the README/docs for background, purpose, and use cases; confirm the version and branch being analyzed; check runtime requirements (dependencies, build method, startup command); quickly verify functionality via sample programs or API calls. Tip: build an overall understanding first rather than diving straight into code.

2. Architecture understanding, from four dimensions: 1) directory structure analysis (list main directories, analyze module division, mark key files); 2) tech-stack identification (languages, frameworks and libraries, architecture patterns); 3) architecture diagramming (draw module-relationship diagrams with draw.io/PlantUML, clarifying call relationships, data flow, and communication); 4) core-flow review (locate the main business flow, identify key phases, draw flowcharts).

3. Deep dive into the core flow: locate the entry file and initialization; trace the core call chain using a debugger or logs, focusing on exception handling, resource release, and locking; analyze data flow (how data passes, mark key data structures, understand the data lifecycle); component analysis (each module's responsibility, interface definitions, whether abstraction layers exist).

4. Focus on details: read complex functions or modules line by line; analyze performance-optimization points (memory pools, caching, async processing); explore edge-case handling (null pointers, timeouts, concurrency conflicts); review error codes and logging.

5. Apply and reflect: summarize design highlights (extensibility, maintainability, high-performance design, design-pattern usage); identify potential problems (memory-leak risks, race conditions, missing configs); propose improvements (how to optimize logic, how to support new features); migrate and apply (transfer the design ideas to your own project, write a technical blog or shareable doc).

🛠️ **适用 AI 工具**：Claude、ChatGPT、Cursor、Copilot、DeepSeek、Gemini