humanerd-drew/opencode-drewgent/site-spec-audit
Claude Code and agent skill by humanerd-drew/opencode-drewgent
- community
Checking that the SKILL.md file is still in the repository.
Install
npx vskill@latest install humanerd-drew/opencode-drewgent/site-spec-auditSKILL.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
어떻게 작동하는가
이 스킬은 두 가지를 조합한다:
- MCP 서버
specification-website—~/.{{AGENT_NAME_LOWER}}/config.yaml의mcp_servers.specification-website에 등록됨 - 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
| Tool | 용도 | |
|---|---|---|
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
2. 카테고리 목록 로드
3. spec 페이지 fetch (status=required)
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.txt존재 + Markdown negotiation 작동 여부
5. 보고서 작성
YOUR_DOMAIN 우선 점검 (Quartz 베이스)
Quartz 빌드 결과를 기준으로 자주 빠지는 항목:
Foundations (14개)
- [ ]
<!doctype html>— Quartz 기본 OK - [ ]
<html lang="ko">— QuartzLocaleconfig OK, 한국어 콘텐츠면ko - [ ]
<meta charset="utf-8">— OK - [ ]
<meta name="viewport">— OK - [ ]
<title>— QuartzTitle.generatedOK - [ ]
<meta name="description">— ⚠️ 페이지별 설정 필요 (frontmatterdescription) - [ ]
<link rel="canonical">— ⚠️ Quartz plugincanonical옵션 - [ ] Favicon + app icons — ⚠️
quartz/static/icon.svg없으면 누락 - [ ]
<meta name="theme-color">— ❌ Quartz 미생성 - [ ]
<meta name="color-scheme">— ❌ Quartz 미생성 - [ ] Open Graph — ⚠️ frontmatter
socialImage,socialDescription필요 - [ ]
<link rel="alternate" type="application/atom+xml">— ❌ RSS feed 발견 안 됨
SEO (13개)
- [ ]
/robots.txt— ❌ Quartz 미생성 - [ ]
/sitemap-index.xml— ⚠️ QuartzenableSiteMap: 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 pluginllmsTxt옵션 확인) - [ ]
/llms-full.txt— ❌ - [ ] MCP server (
/mcp) — ❌ (Cloudflare Pages는 serverless, Workers 필요) - [ ]
Accept: text/markdowncontent negotiation — ❌ - [ ] Content-Signals header — ❌
cron 자동화 (H3)
jobs.json의 site-spec-audit-weekly cron job이 매주 일요일 04:00 KST에 실행:
- Quartz 빌드
- YOUR_DOMAIN 라이브 URL에 spec check 적용
- 보고서를
P6-prefrontal/incidents/site-audit-{date}.md로 저장 - 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
- SKILL.md in the repositoryCheckingskills/site-spec-audit/SKILL.md
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
[](https://verified-skill.com/skills/humanerd-drew/opencode-drewgent/site-spec-audit)