freemcplab / 디렉터리 / 플레이그라운드 / SkillForge

● 커뮤니티 tripleyak ⚡ 바로 사용

SkillForge

Name: SkillForge MCP Server
Author: tripleyak

Phase-based skill router + creator — triages each request, analyzes via 11 thinking lenses, generates XML specs, and runs multi-agent approval before shipping.

SkillForge is a meta-skill for Claude Code and Codex that decides whether to USE an existing skill, IMPROVE one, CREATE new, or COMPOSE from multiple. It runs 4 phases: triage → deep 11-lens analysis → XML spec → generation with multi-agent synthesis. v5.0 streamlined the SKILL.md to 313 lines with lazy-loaded references so the context stays lean.

▶ 데모 보기 📦 설치 💡 사용 사례

왜 쓰나요

핵심 기능

Phase 0 triage: USE_EXISTING / IMPROVE_EXISTING / CREATE_NEW / COMPOSE
Phase 1 deep analysis via 11 thinking lenses before any generation
Phase 2-3 XML-spec-driven generation with iteration
Phase 4 multi-agent synthesis panel must unanimously approve
Lean SKILL.md (313 lines) with lazy-loaded references

라이브 데모

실제 사용 모습

skillforge-skill.replay ▶ 준비됨

0/0

설치

클라이언트 선택

~/Library/Application Support/Claude/claude_desktop_config.json · Windows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "skillforge-skill": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/tripleyak/SkillForge",
        "~/.claude/skills/SkillForge"
      ],
      "_inferred": true
    }
  }
}

Claude Desktop → Settings → Developer → Edit Config 열기. 저장 후 앱 재시작.

~/.cursor/mcp.json · .cursor/mcp.json

{
  "mcpServers": {
    "skillforge-skill": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/tripleyak/SkillForge",
        "~/.claude/skills/SkillForge"
      ],
      "_inferred": true
    }
  }
}

Cursor는 Claude Desktop과 동일한 mcpServers 스키마 사용. 프로젝트 설정이 전역보다 우선.

VS Code → Cline → MCP Servers → Edit

{
  "mcpServers": {
    "skillforge-skill": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/tripleyak/SkillForge",
        "~/.claude/skills/SkillForge"
      ],
      "_inferred": true
    }
  }
}

Cline 사이드바의 MCP Servers 아이콘 클릭 후 "Edit Configuration" 선택.

~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "skillforge-skill": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/tripleyak/SkillForge",
        "~/.claude/skills/SkillForge"
      ],
      "_inferred": true
    }
  }
}

Claude Desktop과 같은 형식. Windsurf 재시작 후 적용.

~/.continue/config.json

{
  "mcpServers": [
    {
      "name": "skillforge-skill",
      "command": "git",
      "args": [
        "clone",
        "https://github.com/tripleyak/SkillForge",
        "~/.claude/skills/SkillForge"
      ]
    }
  ]
}

Continue는 맵이 아닌 서버 오브젝트 배열 사용.

~/.config/zed/settings.json

{
  "context_servers": {
    "skillforge-skill": {
      "command": {
        "path": "git",
        "args": [
          "clone",
          "https://github.com/tripleyak/SkillForge",
          "~/.claude/skills/SkillForge"
        ]
      }
    }
  }
}

context_servers에 추가. 저장 시 Zed가 핫 리로드.

claude mcp add skillforge-skill -- git clone https://github.com/tripleyak/SkillForge ~/.claude/skills/SkillForge

한 줄 명령. claude mcp list로 확인, claude mcp remove로 제거.

사용 사례

실전 활용법: SkillForge

Find the right existing skill instead of reinventing one

👤 Anyone with a big skill library and no memory of what's installed ⏱ ~5 min beginner

언제 쓸까: You have a task and aren't sure if something already covers it.

흐름

Describe the task

Use skillforge. I need to turn our docs site into a searchable knowledge base.✓ 복사됨

→ Triage returns USE_EXISTING with a recommendation (e.g. skill-seekers)
Confirm and hand off

Good — run that skill.✓ 복사됨

→ Routes to the recommended skill

결과: You use what's already there rather than writing a new skill.

함정

Triage is over-eager to create new — Explicitly ask 'prefer USE_EXISTING unless match score < 70%'

함께 쓰기: skill-seekers-skill

Create a new skill with engineered rigor

👤 Skill authors who want quality, not vibes ⏱ ~30 min intermediate

언제 쓸까: Triage returned CREATE_NEW and you want the full 4-phase generation.

흐름

Approve triage

skillforge — CREATE_NEW for a skill that drafts ADRs from design discussions.✓ 복사됨

→ Phase 1 analysis begins with 11 lenses
Review the spec

Show me the XML spec before generation.✓ 복사됨

→ Structured spec you can critique
Run synthesis

Generate, then run the multi-agent panel.✓ 복사됨

→ Panel verdict with unanimous approval or concrete objections

결과: A production-quality SKILL.md ready to install.

함정

Panel stuck in endless loop — Cap iterations — SkillForge supports an iteration limit flag

함께 쓰기: skill-seekers-skill

조합

다른 MCP와 조합해 10배 효율

skillforge-skill + skill-seekers-skill

Use SkillForge to decide, then Skill Seekers to actually ingest sources for CREATE_NEW

skillforge triage, then if CREATE_NEW, use skill-seekers to pull in the docs.✓ 복사됨

skillforge-skill + filesystem

Skill output goes straight into ~/.claude/skills/

Write the generated SKILL.md and references to ~/.claude/skills/<name>/.✓ 복사됨

도구

이 MCP가 노출하는 것

도구	입력	언제 호출
Phase 0 triage	task description + skill library	Start of every request
Phase 1 deep analysis (11 lenses)	task	After triage, before generation
Phase 2-3 XML spec + generation	analysis	CREATE_NEW or IMPROVE_EXISTING
Phase 4 multi-agent synthesis	draft	Final quality gate

비용 및 제한

운영 비용

API 쿼터: None
호출당 토큰: Full 4-phase generation is token-heavy (10-30k)
금액: Free — skill is local
팁: Use triage-only for simple routing; invoke full pipeline only for real skill authoring.

보안

권한, 시크릿, 파급범위

자격 증명 저장: No credentials — skill is prompts

데이터 외부 송신: None from the skill itself

문제 해결

자주 발생하는 오류와 해결

Triage picks wrong verdict

Provide the current skill library list so triage can match properly.

Multi-agent panel keeps rejecting

Check the objections — often a missing section in the spec. Fix once, re-run.

Generated skill feels generic

Give more source material in Phase 1; skill quality follows source quality.

대안

SkillForge 다른 것과 비교

대안	언제 쓰나	단점/장점
skill-seekers-skill	You want document-ingest-first rather than analysis-first	Seekers is heavier on source processing; SkillForge is heavier on engineering rigor
Manual SKILL.md authoring	The skill is small and you have a clear spec	Fastest for trivial skills