vibe-coding-architect

Guides architecture selection and project planning before AI-assisted coding begins.

Updated Aug 9, 2026
One-click install
npx skills add https://github.com/stbui/trade-crawlee --skill vibe-coding-architect-stbui
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: vibe-coding-architect
Source: https://github.com/stbui/trade-crawlee/tree/main/.claude/skills/vibe-coding-architect
Command: npx skills add https://github.com/stbui/trade-crawlee --skill vibe-coding-architect-stbui

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Starting an AI-assisted coding project without a clear architecture leads to messy structure, inconsistent tech stacks, and unmaintainable code. This Skill enforces a structured methodology that locks in architecture decisions before any code is written. ## Core Features & Use Cases - Requirements Clarification: Asks 6 mandatory questions covering product form, runtime environment, project scale, data storage, frontend/backend split, and extensibility expectations. - Architecture Comparison: Presents 2-3 architecture options (single-file, MVC, DDD, microservices) in a comparison table with a recommendation and rationale. - Task Decomposition & Prompt Generation: Breaks the project into independently verifiable modules and produces a 10-part architecture prompt including constraints and acceptance criteria. - Use Case: When a user says "help me build a task management web app", the Skill first clarifies requirements, compares architecture options, decomposes the work into modules, and only then generates a coding prompt for step-by-step implementation. ## Quick Start Ask the assistant to help you plan the architecture for a new project before writing any code, and answer its clarifying questions about product form, environment, and scale.

Frequently Asked Questions about vibe-coding-architect

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I choose an architecture before starting AI-assisted coding?▼

Answer six questions about product form, runtime environment, project scale, data storage, frontend/backend split, and extensibility. Then compare 2-3 architecture options such as single-file, MVC, or DDD in a table and pick one with a documented rationale before writing code.

What questions should I answer before starting a new software project?▼

Clarify the product form (web, app, CLI, agent), runtime environment (local or cloud), project scale (prototype, MVP, or full product), data persistence needs, frontend/backend division, and future extensibility expectations.

MVC vs DDD vs microservices for a small project?▼

Single-file architecture suits prototypes under 300 lines, MVC fits general CRUD and mid-size apps, lightweight DDD suits complex business logic maintained long-term, and microservices fit large teams with independent high-availability modules.

When should I skip architecture planning and just write code?▼

Skip the full planning workflow for bug fixes, small UI or CSS adjustments, and minor features added to an existing project. The methodology is designed for starting new projects from scratch, not incremental maintenance.

Why does AI-generated code become messy in large projects?▼

AI handles large code volumes but struggles with unclear structure. Without explicit architecture constraints, module decomposition, and stated limitations in the prompt, generated code drifts in style, naming, and folder organization across iterations.