From smart-blog-skills
Analyzes blog article quality with a 100-point scoring system across 5 categories: content quality, SEO, E-E-A-T signals, technical elements, and AI readiness. Includes AI content detection (29 humanizer patterns, burstiness, trigger words), PageSpeed integration, and PDF report export.
How this skill is triggered — by the user, by Claude, or both
Slash command
/smart-blog-skills:analyze <檔案或目錄> [--pdf]<檔案或目錄> [--pdf]The summary Claude sees in its skill listing — used to decide when to auto-load this skill
1. 讀取指定的文章檔案
讀取 skills/blog/references/content-rules.md 的評分標準,對文章執行以下分析:
讀取 skills/blog/references/humanizer-patterns.md,掃描文章中出現的 AI 寫作模式:
注意: humanizer 模式與下方觸發詞掃描有部分重疊(模式 #7 ⊃ 觸發詞列表),不重複扣分。
根據 content-rules.md 的 100 分標準打分:
| 類別 | 滿分 |
|---|---|
| 內容品質 | 30 |
| SEO 優化 | 25 |
| E-E-A-T 訊號 | 15 |
| 技術元素 | 15 |
| AI 引用就緒度 | 15 |
## 品質分析報告:[檔案名稱]
### 總分:[N]/100 — [等級]
| 類別 | 得分 | 滿分 |
|------|------|------|
| 內容品質 | [N] | 30 |
| SEO 優化 | [N] | 25 |
| E-E-A-T 訊號 | [N] | 15 |
| 技術元素 | [N] | 15 |
| AI 引用就緒度 | [N] | 15 |
### AI 內容偵測
- 句長爆發性:[數值]([自然/邊界/AI風險])
- AI 觸發詞:[N]/千字([正常/注意/高風險])
- 被動語態:[N]%([正常/注意/高風險])
### 問題清單(依嚴重度排序)
#### 🔴 致命
[列出致命問題]
#### 🟡 高優先
[列出高優先問題]
#### 🟠 中優先
[列出中優先問題]
### 改善建議
[前 3 個最有影響力的改善行動]
如果文章的 frontmatter 包含 url 或 canonical 欄位:
~/.config/smart-blog-skills/config.json 是否存在如果使用者加了 --pdf flag:
{filename}-analysis.pdf需求: Python 3.11+ 和 WeasyPrint (pip install weasyprint)
/c/Users/EricHsu/AppData/Local/Programs/Python/Python313/python.exe scripts/pdf_report.py --input report.md --output report.pdf
如果 Python 或 WeasyPrint 未安裝,提示使用者安裝方式,繼續輸出 Markdown 報告。
分析完成後,如果 docs/monitor/ 目錄存在:
docs/monitor/snapshots/{today}/{slug}.jsondocs/monitor/index.json如果 docs/monitor/ 不存在,跳過此步驟。
如果使用者提供目錄路徑:
Glob 找出所有 .md / .mdx / .html 檔案docs/monitor/ 存在,自動建立批次快照npx claudepluginhub rainday/smart-blog-skills --plugin smart-blog-skillsAudits blog posts with a 0–100 score across content quality, SEO, E-E-A-T, technical elements, and AI citation readiness. Includes AI content detection, batch analysis, and prioritized fix recommendations for local files or URLs.
Scans all blog files for quality scores, orphan pages, topic cannibalization, stale content, and AI citation readiness. Spawns parallel subagents for comprehensive analysis and produces a prioritized action queue.
Analyzes content quality, readability, thin content risk, and E-E-A-T signals for SEO. Useful for content-focused audits and AI citation readiness checks.