AgentSkillsCN

surfer-seo-optimizer

基于数据驱动的内容优化助手，通过逆向分析排名靠前的页面，提供切实可行的内容指导、评分标准以及关键词推荐。（仿照 SurferSEO 模式打造）

SKILL.md
--- frontmatter
name: surfer-seo-optimizer
description: Data-driven content optimization agent that reverse-engineers top ranking pages to provide actionable content guidelines, scoring, and keyword recommendations. (Modeled after SurferSEO)
version: 1.0.0

SurferSEO Optimizer Skill

🧠 Core Philosophy

This skill treats content as data. It does not guess. It uses the "Reverse Engineering" method to determine what Google rewards for a specific keyword. Instead of generic advice ("write good content"), it provides specific, quantifiable targets (e.g., "Hit 2,100 words", "Use 'startup costs' 3 times").

⚡ Capabilities

1. SERP Analysis (analyze)

Goal: Determine the "Perfect Content Structure" for a target keyword. Protocol:

  1. •Search: Use search_web to find the top 5 ranking articles for the keyword.
  2. •Extract: For each article, determine:
    • •Word Count (approximate)
    • •Heading Structure (H2/H3 usage)
    • •Key Topics/Entities covered
  3. •Synthesize: Create a "Content Brief" defining:
    • •Target Word Count (Average of Top 5 + 10%)
    • •Required Sections (H2s that appear in most competitors)
    • •"NLP Keywords" (Terms that appear frequently across top pages)

2. Content Audit (audit)

Goal: Score a local MDX/MD file against the "Perfect Structure". Protocol:

  1. •Read: Read the local file content.
  2. •Measure: Run python .agent/skills/surfer-seo-optimizer/scripts/score_content.py (or internal logic) to calculate:
    • •Current Word Count
    • •Keyword Density
    • •Heading Count
  3. •Grade: Assign a score (0-100) based on closeness to targets.
    • •Green: Within 10% of target.
    • •Yellow: Within 25% of target.
    • •Red: Missed target significantly.

3. NLP Optimization (optimize)

Goal: Rewrite content to improve the score without sacrificing readability. Protocol:

  1. •Identify a paragraph or section that is "thin" or missing keywords.
  2. •Rewrite it to naturally include the missing terms.
  3. •Constraint: Do not "keyword stuff". Readability > SEO.

📝 Usage Example (Workflow)

markdown
User: /surfer analyze "startup valuation methods"
Agent: [Searches Web] -> [Generates Brief: 2500 words, Keywords: 'DCF', 'Berkus Method'...]

User: /surfer audit content/startup-valuation.mdx "startup valuation methods"
Agent: [Reads File] -> [Runs Scoring] -> "Your Score: 62/100. You are missing 'Berkus Method'."

User: /surfer optimize "Add the section about Berkus Method"
Agent: [Writes Content] -> "Added section..."

📂 Resources

  • •scripts/score_content.py: CLI tool for text statistics.