harness-create-architecture

Create architecture decision documentation through a guided collaborative workflow.

1|Updated May 23, 2026
One-click install
npx skills add https://github.com/baobao0303/harness --skill harness-create-architecture
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: harness-create-architecture
Source: https://github.com/baobao0303/harness/tree/main/.agents/skills/harness-create-architecture
Command: npx skills add https://github.com/baobao0303/harness --skill harness-create-architecture

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) and assets (resource) components.

What problem does it solve?

This Skill prevents AI agents from producing inconsistent or conflicting architecture by guiding collaborative, step-by-step discovery and documentation.

Core Features & Use Cases

  • Architecture decision orchestration: drives a structured workflow from project context through decisions, patterns, structure, and validation.
  • Multi-agent consistency guardrails: captures implementation patterns and enforces conflict-preventing rules across agents.
  • Document-first collaboration: builds an append-only architecture document with tracked state and loaded facts from project docs.

Quick Start

Ask an AI agent to run the harness-create-architecture workflow when you say: "Let's create architecture for this project using our PRD and project docs."

Frequently Asked Questions about harness-create-architecture

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

FAQPage Schema
How do I create consistent architecture decision documentation for AI agents?

Architecture decision records are structured documents that capture project context, chosen patterns, and implementation boundaries. They are needed when multiple AI agents collaborate on a project to prevent conflicting design choices and ensure version-aware consistency.

How do I generate project scaffolding from a PRD and existing project docs?

You generate project scaffolding by instructing an AI agent to load project context from configured sources like a PRD and project docs. The workflow appends artifacts deterministically by steps, blocking progression without explicit user confirmation.

What is the best way to prevent conflicting architecture decisions across multiple AI agents?

The best way to prevent conflicting architecture decisions is using a document-first collaboration workflow with multi-agent consistency guardrails. It captures implementation patterns and enforces conflict-preventing rules across agents within an append-only document.

Do I need explicit user confirmation for each step of a collaborative architecture workflow?

Yes, explicit user confirmation is required for each step of the collaborative architecture workflow. The workflow blocks progression until you confirm, ensuring structured, version-aware decisions align with your project boundaries and loaded facts.

Can I use existing project documentation to drive architecture discovery and validation?

Yes, you can use existing project documentation to drive architecture discovery. The workflow loads project context directly from configured sources and builds an append-only architecture document with tracked state and loaded facts.