로그인
miles-knowbl

portability

by @miles-knowbl in AI & LLM
1
0
# Install this skill:
npx skills add miles-knowbl/orchestrator --skill "portability"

Install specific skill from multi-skill repository

# Description

Assess and document skill transferability across projects and domains, identifying required adaptations and context dependencies.

# SKILL.md

name: portability
description: "Assess and document skill transferability across projects and domains, identifying required adaptations and context dependencies."
phase: COMPLETE
category: meta
version: "1.0.0"
depends_on: ["skill-verifier"]
tags: [meta, portability, reusability, transfer, assessment]

Portability

Assess skill transferability across projects and domains.

When to Use

  • After skill creation — Evaluate how portable a new skill is
  • Cross-project reuse — Adapting a skill for a different project
  • Skill library curation — Deciding what belongs in the shared library
  • When you say: "is this reusable", "port this skill", "how transferable"

Required Deliverables

Deliverable Location Condition
Portability assessment Skill's SKILL.md (section) Always

Core Concept

Portability assessment answers: "Can this skill be applied to a different project without modification?"

Fully Portable ←──────────────────────→ Completely Domain-Specific
  (works anywhere)                        (only works here)

Portability Levels

Level Description Example
Universal Works for any project code-review, test-generation
Category-portable Works within a category (web, CLI, API) frontend-design, infra-docker
Stack-portable Works within a tech stack (Node.js, Python) infra-database (Drizzle-specific)
Domain-specific Tied to a specific domain or project Custom business logic skills

Assessment Criteria

Criterion Portable Not Portable
Language references Generic or multi-language Single language only
Framework dependencies Framework-agnostic Tied to specific framework
Domain concepts General engineering Business-specific terminology
File paths Configurable or conventional Hardcoded paths
External dependencies Standard tools Proprietary or rare tools

Adaptation Catalog

When porting a skill, common adaptations include:

Adaptation Description
Language swap Replace code examples with target language
Framework swap Replace framework patterns (Express → Fastify)
Convention mapping Map naming conventions to target project
Tool substitution Replace tools (Drizzle → Prisma, vitest → jest)
Scope adjustment Expand or narrow scope for target context

Checklist

  • [ ] Portability level assessed (universal → domain-specific)
  • [ ] Context dependencies identified
  • [ ] Required adaptations documented
  • [ ] Generic vs. specific sections clearly separated
  • [ ] Examples provided in multiple languages/frameworks (if portable)

Relationship to Other Skills

Skill Relationship
skill-verifier Verifier checks skill quality; portability checks transferability
skill-design Design decisions affect portability
retrospective Portability insights feed into future skill design

Key Principles

Generic by default. Write skills for the general case; specialize only when necessary.

Separate the portable from the specific. Keep domain-specific content in references, not the core skill.

Adaptation over duplication. One portable skill with adaptations beats five copies.

Document the dependencies. Make implicit context explicit so porters know what to change.

# Related Skills

# Supported AI Coding Agents

This skill is compatible with the SKILL.md standard and works with all major AI coding agents:

⚡ Amp 🚀 Antigravity 🤖 Claude Code 🦀 Clawdbot 📝 Codex ▶️ Cursor 🤖 Droid 💎 Gemini CLI 🐙 GitHub Copilot 🪿 Goose 📊 Kilo Code 🔧 Kiro CLI 💻 OpenCode 🦘 Roo Code 🌲 Trae 🏄 Windsurf

Learn more about the SKILL.md standard and how to use these skills with your preferred AI coding agent.