ArchGuard backend

Maven Central Docker Image Version

ArchGuard is an architecture governance tool that can analysis architecture in container, component, code level, database, create architecture fitness functions, and test for architecture rules.

Chinese: ArchGuard 是一个针对于微服务（分布式场景）下的架构工作台/治理工具。它可以帮助架构师、开发人员进行架构自助，自定义架构的洞察、分析系统间的远程服务依赖情况、数据库依赖、API 依赖等。并根据一些架构治理模型，对现有系统提出改进建议。

Document: https://archguard.org/
Roadmap: Roadmap
Contribute: Contribute to Archguard
SubProjects:
- ArchGuard Frontend
- Chapi source code analysis
- ArchGuard Co-mate an AI-powered architecture copilot, design and governance tools.
Architecture:

Architecture

You can use:

ArchGuard Library in Maven Central to integrate with your backend.
ArchGuard Scanner CLI to scan your codebase, upload to your backend or ArchGuard backend.
ArchGuard Gradle Plugin to scan your codebase in CI/CD.
ArchGuard Web is the backend of ArchGuard, it provides a RESTful API for frontend.

特性（Features）：

设计态
- 架构设计、分析与治理 DSL
- Feakin: https://github.com/feakin/fklang
开发态
- 架构扫描
  - 扫描配置
  - 插件化规则定制
  - 规则化治理：Code Smell, Test Code Smell, SQL Smell, API Smell, Documentation Smell, etc.
- 架构可视化
  - 基于 C4 模型的可视化分析
    - 上下文：API 服务地图（API 生产者支持语言：Java、Kotlin、C#，API 消费者支持语言：TypeScript/JavaScript、Kotlin、Java 等）
    - 容器分析。数据库地图（支持 MyBatis、JDBI、JPA）
    - 组件分析
    - 代码分析：支持级别模块、包、类、方法四个级别。
  - 高级分析 + 可视化
    - 系统不稳定性模块分析。
    - 容器间：精准测试/变化分析
- 架构指标（单体DONE，分布式DOING）
  - 体量维度：过大的组件
  - 耦合维度：枢纽组件，过深调用，循环依赖
  - 内聚维度：霰弹式修改
  - 冗余维度：冗余元素，过度泛化
  - 质量维度：测试保护
- 代码分析
  - CLOCO：代码复杂度 #79
  - SCA 分析
  - OpenAPI 分析
  - Architecture analysis
运行态
- APM（TODO）
架构工作台

Features：

Design State
- Architecture Design, Analysis and Governance DSL
- Feakin: https://github.com/feakin/fklang
Development state
- Schema scan
  - Scan configuration
  - Plug-in rule customization
  - Rule-based governance: Code Smell, Test Code Smell, SQL Smell, API Smell, Documentation Smell, etc.
- Architecture visualization
  - Visual analysis based on C4 model
    - Context: API service map (API producer supported languages: Java, Kotlin, C#, API consumer supported languages: TypeScript/JavaScript, Kotlin, Java, etc.)
    - Container analysis. Database map (support MyBatis, JDBI, JPA)
    - Component analysis
    - Code analysis: supports four levels of modules, packages, classes, and methods.
  - Advanced Analysis + Visualization
    - System instability module analysis.
    - Between containers: precise testing/variation analysis
- Architecture metrics (single DONE, distributed DOING)
  - Volume dimension: oversize components
  - Coupling dimension: hub components, too deep calls, circular dependencies
  - Cohesive Dimension: Shotgun Modification
  - Redundant dimensions: redundant elements, overgeneralization
  - Quality dimension: test protection
  - Continuous Integration
- External analysis
  - CLOCO: Code Complexity #79
  - SCA analysis
  - OpenAPI analysis
  - Architecture analysis
Running state
- APM (TODO)
Architecture Workbench

Screenshots:

Languages parse by Chapi

| Features/Languages | Java | Python | Go | Kotlin | TypeScript | C | C# | Scala | C++ | |---------------------|------|--------|----|--------|------------|----|----|-------|-----| | http api decl | ✅ | 🆕 | 🆕 | ✅ | ✅ | 🆕 | ✅ | 🆕 | 🆕 | | syntax parse | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | 🆕 | | function call | ✅ | 🆕 | ✅ | ✅ | ✅ | | | | | | arch/package | ✅ | | | ✅ | ✅ | 🆕 | ✅ | ✅ | | | real world validate | ✅ | | | | ✅ | | | | |

Custom Backend

case example:

https://github.com/unit-mesh/co-unit by Rust language

use Scanner CLI you can customize your backend. For more detail, see in: ArchGuardHttpClient

HTTP examples:

POST http://127.0.0.1:8765/scanner/:systemId/reporting/class-items

POST http://127.0.0.1:8765/scanner/:systemId/reporting/openapi

POST http://127.0.0.1:8765/scanner/:systemId/reporting/container-services

POST http://127.0.0.1:8765/scanner/:systemId/reporting/datamap-relations

...

MCP Server (Model Context Protocol)

🚀 New Feature Proposal: Expose ArchGuard's architecture linter as an MCP server for AI-assisted development!

The MCP server would enable AI assistants (Claude, ChatGPT, etc.) and development tools to leverage ArchGuard's linting capabilities through a lightweight, standardized interface.

Key Benefits:

⚡ Lightweight: <100MB memory, <2s startup (vs. full server)
🤖 AI Integration: Works with Claude, ChatGPT, GitHub Copilot
🔌 Standard Protocol: Compatible with any MCP-enabled tool
🎯 Stateless: Quick analysis without database overhead

Documentation:

Detailed Proposal - Full specification and architecture
GitHub Issue Template - Ready to submit feature request
Implementation Guide - Step-by-step development guide

This would complement the existing server and CLI, adding a modern AI-friendly interface for architectural governance.

Chat

关注我们：

欢迎加入我们：

（PS：如果群满，请添加微信 phodal02，并注明 ArchGuard）

Thanks

JetBrains support:

License

This code is distributed under the MIT license. See LICENSE in this directory.

Archguard

Install / Use

README