Can I use these instructions for human developers?

Yes, the skill generates 'Human Instructions' which serve as executable documentation with copy-pasteable commands, allowing developers to manually verify or complete tasks.

Does it support OpenCode integration?

Yes, it includes comprehensive protocols, setup guides, and migration paths for dual-mode execution using both Claude and OpenCode agents.

What is the primary role of an Architect Agent?

The Architect Agent focuses on high-level planning, defining technical requirements, and delegating specific tasks to Code Agents to ensure separation of concerns between design and implementation.

How long does the workspace setup take?

With the included template-based bash scripts, you can initialize both Architect and Code Agent workspaces in less than five minutes.

How does the grading system work?

The skill uses a 6-category grading rubric to evaluate the work completed by Code Agents, ensuring that implementations meet architectural standards and project requirements.

Architect Agent Workflow

Name: Architect Agent Workflow
Author: SpillwaveSolutions

bySpillwaveSolutions

•

生产力与工作流

Manages the end-to-end orchestration of AI agent hierarchies by coordinating planning, instruction delegation, and quality evaluation.

The Architect Agent Skill introduces a sophisticated multi-agent orchestration layer for Claude Code, enabling users to separate high-level architectural planning from tactical code implementation. By establishing distinct workspaces for Architect and Code agents, this skill provides a structured framework for generating executable instructions, initializing standardized project environments, and performing rigorous automated grading of agent output against predefined rubrics. It is particularly useful for managing complex software projects where maintaining architectural integrity and quality control across multiple AI-driven development tasks is critical.

主要功能

01Multi-agent workspace orchestration and automated setup

02Structured instruction generation for code-focused agents

03Cross-workspace permission and hook configuration

04Standardized grading rubrics for evaluating implementation quality

0515 GitHub stars

06OpenCode integration for dual-mode agent execution

使用场景

01Implementing a quality assurance loop where AI-generated code is automatically graded against project standards

02Establishing standardized, repeatable development environments for distributed AI agent teams

03Delegating complex software tasks from a lead Architect agent to specialized Code agents

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add spillwavesolutions/agent_rulez architect-agent

For use in Claude.ai and ChatGPT

Download Skill