design

Analyze codebases and research findings to produce detailed architecture plans.

1|1|Updated Feb 13, 2026
One-click install
npx skills add https://github.com/laststance/skills --skill design-laststance
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: design
Source: https://github.com/laststance/skills/tree/main/skills/design
Command: npx skills add https://github.com/laststance/skills --skill design-laststance

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Design of complex systems often lacks a rigorous, repeatable process; this Skill provides an architecture-driven plan creation workflow to produce exhaustive, implementable designs.

Core Features & Use Cases

  • Phase-based research, planning, design, validation, and documentation.
  • Multi-agent review loops (architecture, completeness, feasibility) to ensure coherence.
  • Generates plan artifacts saved to docs/plans with traceable references.

Quick Start

Provide a target feature or system and any constraints to begin an architecture design plan.

Frequently Asked Questions about design

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

FAQPage Schema
How do I create an implementation plan for a complex system architecture?

To create an implementation plan for system architecture, use a phase-based workflow that conducts research, designs the architecture, validates completeness, and generates an export-ready plan artifact. This ensures your system design is robust and traceable to original requirements.

What is a multi-agent review loop in architecture design?

A multi-agent review loop in architecture design is a validation mechanism that coordinates separate agents to evaluate architecture coherence, feature completeness, and technical feasibility, ensuring the final plan is exhaustive and implementable without gaps.

How do I ensure my architecture design plan is fully traceable?

To ensure your architecture design plan is fully traceable, generate plan artifacts that include annotated references linking design decisions back to codebase analysis and research findings. This guarantees every architectural choice is backed by documented evidence.

Can I use this approach to plan both individual features and entire systems?

Yes, you can use this architecture design approach to plan both individual features and entire systems. It scales across different scopes by applying the same exhaustive research, multi-phase design, and formal review loops to any target component.

What is the best way to document robust architecture plans for export?

The best way to document robust architecture plans for export is to save them as structured artifacts within a designated directory like docs/plans. This standardizes the output, making the implementation designs ready for downstream development and team distribution.