humanerd-drew/opencode-drewgent/site-spec-audit

Claude Code and agent skill by humanerd-drew/opencode-drewgent

Trust tier 3
  • community
Source check pending

Checking that the SKILL.md file is still in the repository.

Install

npx vskill@latest install humanerd-drew/opencode-drewgent/site-spec-audit

SKILL.md

The same SKILL.md was published earlier as humanerd-drew/opencode-drewgent/site-spec-audit.


name: site-spec-audit title: site-spec-audit — YOUR_DOMAIN을 The Website Spec으로 감사 description: Audit any URL (default: YOUR_DOMAIN) against The Website Specification (specification.website). Uses the specification-website MCP server (search, list_topics, get_topic, get_checklist, get_categories). Reports required items first, then recommended, with primary sources cited. Use when the user asks to audit/check/validate YOUR_DOMAIN, when a new site is deployed, or when planning a new feature that touches foundations/seo/accessibility/security/agent-readiness. type: skill space: growth tags: [skill, audit, web-standards, YOUR_SITE, agent-readiness] created: 2026-06-01 updated: 2026-06-01 links:

  • "[[skills/specification-website]]"
  • "[[skills/YOUR_SITE]]"
  • "[[skills/seo-audit]]"
  • "[[skills/filesystem-truth-audit]]"
  • "[[@action/skills/SKILL-INDEX]]"
  • "[[@identity/brain/rules]]"---

site-spec-audit

YOUR_DOMAIN (or any URL) 을 The Website Specification으로 감사하는 스킬. 기본 대상은 https://YOUR_DOMAIN. 다른 URL은 인자로 전달.

사용법

/site-spec-audit # YOUR_DOMAIN 감사
/site-spec-audit https://example.com # 다른 URL 감사
/site-spec-audit --category=agent-readiness # 카테고리 한정 감사
/site-spec-audit --status=required # status=required만
/site-spec-audit --status=recommended # status=recommended만

어떻게 작동하는가

이 스킬은 두 가지를 조합한다:

  1. MCP 서버 specification-website — ~/.{{AGENT_NAME_LOWER}}/config.yaml의 mcp_servers.specification-website에 등록됨
  2. HTTP fallback — https://specification.website/.well-known/agent-skills/specification-website/SKILL.md의 워크플로우대로 직접 fetch

MCP 도구 우선. MCP가 안 되면 llms.txt / 직접 fetch로 fallback.

MCP 도구

Tool용도
get_categories()10개 카테고리 + 카운트 + 요약
list_topics({ category, status, limit })필터된 인덱스
get_topic({ slug })단일 spec 페이지 풀 마크다운
get_checklist({ category, status })체크리스트 형식
search(query, limit)전문 검색

Prompt: audit_url(url, focus?) — 대상 URL에 대한 감사 플랜 생성.

감사 워크플로우 (5단계)

1. 스코프 결정
url = 인자 또는 기본값 "https://YOUR_DOMAIN"
focus = 인자의 --category 또는 None (전체 10 카테고리)
status_filter = 인자의 --status 또는 "required" 먼저
2. 카테고리 목록 로드
python
mcp.call("get_categories") # 10개 카테고리 + 카운트
3. spec 페이지 fetch (status=required)
python
for category in categories:
mcp.call("list_topics", {"category": category, "status": "required"})
for topic in topics:
spec = mcp.call("get_topic", {"slug": topic.slug})
# spec의 "How to verify" 섹션 → 체크리스트 항목으로 변환
4. URL 실제 검증

각 spec 항목에 대해 YOUR_DOMAIN에서 검증:

  • HTTP fetch (web_search / browser_navigate) — og:title, <meta charset>, <title> 등 HTML head 검증
  • well-known fetch — /.well-known/security.txt, /.well-known/agent-skills/, /robots.txt, /sitemap-index.xml
  • Accept: text/markdown — /llms.txt, /llms-full.txt 존재 + Markdown negotiation 작동 여부
5. 보고서 작성
# YOUR_DOMAIN 감사 보고서 — YYYY-MM-DD
## 통과: N개
- ✅ <title> element present
- ✅ <meta charset="UTF-8"> set
- ...
## 누락 (required): M개
- ❌ <meta name="theme-color"> missing
- ❌ /robots.txt returns 404
- ❌ /sitemap-index.xml missing (Quartz 기본 생성 안 함)
- ...
## recommended 미흡: K개
- ⚠️ /llms.txt recommended but not present
- ⚠️ Open Graph tags present but og:image missing
- ...
## 출처
각 spec 항목의 `sources` 필드에서 1차 표준 인용
(W3C, WHATWG, IETF RFC, WCAG, MDN)

YOUR_DOMAIN 우선 점검 (Quartz 베이스)

Quartz 빌드 결과를 기준으로 자주 빠지는 항목:

Foundations (14개)
  • [ ] <!doctype html> — Quartz 기본 OK
  • [ ] <html lang="ko"> — Quartz Locale config OK, 한국어 콘텐츠면 ko
  • [ ] <meta charset="utf-8"> — OK
  • [ ] <meta name="viewport"> — OK
  • [ ] <title> — Quartz Title.generated OK
  • [ ] <meta name="description"> — ⚠️ 페이지별 설정 필요 (frontmatter description)
  • [ ] <link rel="canonical"> — ⚠️ Quartz plugin canonical 옵션
  • [ ] Favicon + app icons — ⚠️ quartz/static/icon.svg 없으면 누락
  • [ ] <meta name="theme-color"> — ❌ Quartz 미생성
  • [ ] <meta name="color-scheme"> — ❌ Quartz 미생성
  • [ ] Open Graph — ⚠️ frontmatter socialImage, socialDescription 필요
  • [ ] <link rel="alternate" type="application/atom+xml"> — ❌ RSS feed 발견 안 됨
SEO (13개)
  • [ ] /robots.txt — ❌ Quartz 미생성
  • [ ] /sitemap-index.xml — ⚠️ Quartz enableSiteMap: true로 가능
  • [ ] <link rel="canonical"> (위와 중복)
  • [ ] Heading hierarchy (h1→h2→h3) — OK (Quartz 컴포넌트가 자동)
  • [ ] JSON-LD structured data — ❌ Quartz 미생성 (직접 emitter 필요)
Well-Known URIs (9개)
  • [ ] /.well-known/security.txt — ❌ (Cloudflare Pages에서 static file로 가능)
  • [ ] /.well-known/agent-skills/ — ❌ (인덱스 + SKILL.md)
  • [ ] /.well-known/api-catalog — ❌
Agent Readiness (18개)
  • [ ] /llms.txt — ❌ (Quartz plugin llmsTxt 옵션 확인)
  • [ ] /llms-full.txt — ❌
  • [ ] MCP server (/mcp) — ❌ (Cloudflare Pages는 serverless, Workers 필요)
  • [ ] Accept: text/markdown content negotiation — ❌
  • [ ] Content-Signals header — ❌

cron 자동화 (H3)

jobs.json의 site-spec-audit-weekly cron job이 매주 일요일 04:00 KST에 실행:

  1. Quartz 빌드
  2. YOUR_DOMAIN 라이브 URL에 spec check 적용
  3. 보고서를 P6-prefrontal/incidents/site-audit-{date}.md로 저장
  4. 회귀 감지 (이전 보고서와 diff)

Sources

  • The Website Specification: https://specification.website/
  • Spec SKILL: https://specification.website/.well-known/agent-skills/specification-website/SKILL.md
  • YOUR_SITE skill: [[skills/YOUR_SITE]]

Security

  • Security scanPassed, 90 of 100Pattern checks on the skill's files.
  • Model reviewPassed

Quality evaluation

  • Task evaluationNot runNo evaluation yet. Run one in Skill Studio.

Source

Popularity

7-day trend score
0

Not refreshed yet.

Works with

Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.

  • Claude Code
  • Cursor
  • GitHub Copilot
  • Windsurf
  • Codex
  • Gemini CLI
  • Cline
  • Amp
  • Roo Code
  • Goose
  • Continue
  • Kilo Code
  • Junie
  • Trae
  • OpenHands
  • Qwen Code
  • AI Agent
  • Tabnine Chat
  • Sourcegraph Cody
  • Amazon Q
  • Firebase Genkit
  • Sweep
  • Devin
  • Phind
  • Mentat
  • Claude Desktop
  • Aider
  • GPT Pilot
  • Plandex
  • bolt.new
  • v0
  • Replit Agent
  • Anse
  • TypingMind
  • BetterChatGPT
  • LibreChat
  • Warp AI
  • Pieces
  • Fig AI
Embed badge
vskill VERIFIED badge
[![vskill VERIFIED](https://verified-skill.com/api/v1/skills/humanerd-drew/opencode-drewgent/site-spec-audit/badge)](https://verified-skill.com/skills/humanerd-drew/opencode-drewgent/site-spec-audit)

Something wrong with this skill or its page? Tell us.