A curated collection of effective prompts for AI coding assistants to enhance your development workflow and productivity.
This repository contains several specialized AI roles designed for different technical tasks:
- Role: Code Archaeologist, Visual Designer & Architecture Theorist
- Purpose: Generate high-quality, high-fidelity core architecture diagrams based on C4 Model and 4+1 View Model
- Key Features:
- Reverse engineer real architecture from source code
- Create pixel-perfect SVG diagrams following strict visual specifications
- Apply advanced architectural theories (C4 Model, 4+1 View Model)
- Role: Dynamic Systems Archaeologist
- Purpose: Create comprehensive sequence diagrams documenting all system interactions
- Key Features:
- Automatically discover and document all system flows
- Generate exhaustive Mermaid sequence diagrams
- Produce a complete "Dynamic Behavior Codex" for the project
- Role: Technical Product Documentation Specialist
- Purpose: Transform scattered technical materials into systematic product documentation
- Key Features:
- Reverse engineer product requirements from code and diagrams
- Create structured PRD with prioritized feature lists
- Maintain glossary and ensure documentation coherence
All roles in this repository depend on specific MCP (Model Context Protocol) tools:
sequentialthinking
(Required by ALL roles) - Enables strategic, step-by-step planning and executioncontext7
(Required by System Architect) - Provides real-time access to architectural theory documentationmcp-feedback-enhanced
(Recommended for all roles) - Enables interactive feedback during execution
-
Install Required MCP Tools:
# Install sequentialthinking (mandatory) # Install context7 (for architecture diagrams) # Install mcp-feedback-enhanced (recommended)
-
Choose Your Role: Navigate to the relevant directory based on your task:
- Architecture diagrams →
generate-diagrams/
- Product documentation →
generate-prd/
- Architecture diagrams →
-
Copy and Use: Copy the role prompt to your AI assistant (e.g., Cursor) and modify according to your specific context
-
Follow the Workflow: Each role has a specific workflow that leverages the MCP tools for optimal results
This project is licensed under the MIT License - see the LICENSE file for details.
一个精心策划的 AI 编程助手提示集合,旨在提升您的开发工作流程和生产力。
本仓库包含多个专门设计用于不同技术任务的 AI 角色:
- 角色:代码考古学家、视觉设计师与架构理论家
- 用途:基于 C4 Model 和 4+1 View Model 生成高质量、高保真的核心架构图
- 主要特性:
- 从源代码反向推导真实架构
- 遵循严格视觉规范创建像素级完美的 SVG 图表
- 应用先进的架构理论(C4 Model、4+1 View Model)
- 角色:动态系统考古学家
- 用途:创建记录所有系统交互的全面时序图
- 主要特性:
- 自动发现并记录所有系统流程
- 生成详尽的 Mermaid 时序图
- 产出项目的完整"动态行为白皮书"
- 角色:技术产品文档专家
- 用途:将零散的技术材料转化为系统化的产品文档
- 主要特性:
- 从代码和图表反向工程产品需求
- 创建带优先级功能列表的结构化 PRD
- 维护术语表并确保文档连贯性
仓库中的所有角色都依赖于特定的 MCP(模型上下文协议)工具:
sequentialthinking
(所有角色必需)- 实现战略性、分步骤的规划和执行context7
(系统架构师必需)- 提供架构理论文档的实时访问mcp-feedback-enhanced
(所有角色推荐)- 在执行过程中启用交互式反馈
-
安装必需的 MCP 工具:
# 安装 sequentialthinking(必需) # 安装 context7(用于架构图) # 安装 mcp-feedback-enhanced(推荐)
-
选择您的角色:根据您的任务导航到相关目录:
- 架构图 →
generate-diagrams/
- 产品文档 →
generate-prd/
- 架构图 →
-
复制并使用:将角色提示复制到您的 AI 助手(如 Cursor)并根据您的特定上下文进行修改
-
遵循工作流程:每个角色都有特定的工作流程,利用 MCP 工具获得最佳结果
该项目采用 MIT 许可证 - 有关详细信息,请参阅 LICENSE 文件。