Archbase — Architecture that AI coding agents can actually follow

I've been building Archbase, an open-source CLI written in Go that helps developers and AI coding agents follow consistent software architecture and coding standards.

AI coding agents can generate code quickly, but keeping that code aligned with a project's architecture, conventions, and dependency boundaries remains a challenge.

Archbase aims to solve this.

It provides:

Architecture Patterns: Reusable structures and conventions for different types of applications.

Architecture Rules: Centralized rules for responsibilities, dependencies, and code organization.

Project Scopes: Context-aware pattern selection for different parts of a codebase.

AI Agent Integration: Support for Cursor, GitHub Copilot, Claude, AGENTS.md, and read-only MCP tools.

Offline-first CLI: Built in Go, with an interactive terminal interface.

The core CLI and its architecture catalog are open source under the MIT license.

I'm interested in connecting with other developers and AI founders working on developer tools, coding agents, and AI-assisted software engineering.

I'd love to hear how others are maintaining architectural consistency in AI-generated code.

GitHub: github.com/EnzoCaetano015/Archbase

Website: archbase.caetanodev.com/