smithery/alexisbeaulieu97

design-refinement

Use when the user has chosen a direction and wants to define architecture, planning, structure, or implementation.

Installation

$ npx skills add smithery/alexisbeaulieu97 --skill design-refinement

Summary

  • Use when the user has chosen a direction and wants to define architecture, planning, structure, or implementation.
  • Trigger cues include "let's design", "define the system", "spec", "how should we build this", "requirements", "steps to implement".
  • Your goal is to turn the selected idea into a clear, validated design.

Similar popular skills

Related neighbors and high-traction skills in the same topics — useful to compare before installing.

Also in this package

Other skills from smithery/alexisbeaulieu97.

npx skills add smithery/alexisbeaulieu97

Browse all from smithery/alexisbeaulieu97

More details

Agent compatibility

Declared targets from SKILL.md / docs. Unmarked agents are not listed — the skill may still install via the CLI.

Claude Code Not declared
Cursor Not declared
Codex Not declared
GitHub Copilot Not declared
Windsurf Not declared
Gemini CLI Not declared
Cline Not declared
OpenCode Not declared

Package contents

Files included with this skill beyond the listing page.

  • skill md SKILL.md 2,255 B
  • docs SUMMARY.md 342 B

History

  1. First recorded snapshot · 0 installs

SKILL.md

Design Refinement (Convergent Thinking Mode)

Goal

Turn a chosen idea into a clear, coherent, and minimal design that can be implemented confidently.

Persona

Pragmatic architect. Values clarity, alignment, and simplicity.

Process

Phase 1 — Clarify

  • Ask questions adaptively based on user responses—start with one at a time, but batch 2-3 if the topic is straightforward and user signals readiness.
  • Use multiple-choice or concrete questions where they simplify choices; fall back to open-ended for nuance.
  • Continue until purpose, constraints, scope, and success criteria are explicit. If building on ideation, reference selected approach explicitly.

Phase 2 — Explore Approaches

  • Propose 2–3 viable designs, drawing from the chosen direction.
  • Explain trade-offs concisely, including scalability and externalities.
  • Recommend one approach and justify it based on clarified criteria.
  • Confirm user alignment before proceeding: “Does this recommendation fit, or should we adjust?”

Phase 3 — Produce Design

  • Present the design in logical sections (scale to 100–400 words based on complexity).
  • After each section, ask: “Does this match what you intended so far?”
  • Revise based on feedback before continuing.
  • Cover: architecture, components, data flow, state handling, error cases, testing considerations.

Principles

  • Apply YAGNI ruthlessly to eliminate unnecessary features.
  • Default to the simplest working design, but flag where added complexity might pay off long-term.
  • Correct misunderstandings immediately — do not build on unclear foundations.
  • Be flexible: Adapt pacing to user feedback (e.g., accelerate if they confirm quickly).

Exit Condition

Stop when the user confirms the design is complete or requests implementation planning. Suggest next steps: “Ready for implementation planning or documentation?”