2026 年用 Claude API 构建自己的 SEO 自动化工具：代码与技巧

npm install @anthropic-ai/sdk

import Anthropic from '@anthropic-ai/sdk'

const client = new Anthropic({ apiKey: process.env.ANTHROPIC_API_KEY })

async function generateContentBrief(keyword: string): Promise<string> {
  const message = await client.messages.create({
    model: 'claude-3-5-sonnet-20241022',
    max_tokens: 2000,
    messages: [{
      role: 'user',
      content: `Create a detailed SEO content brief for the keyword: "${keyword}"

Include:
1. Target audience and search intent
2. Recommended H1, H2, and H3 structure (10+ headings)
3. Key subtopics to cover for comprehensive coverage
4. FAQ section (6 questions people commonly ask)
5. Suggested internal linking targets
6. Meta title (under 60 chars) and meta description (under 160 chars)
7. Target word count range

Format as a structured document I can use to write the post.`
    }]
  })

  return message.content[0].type === 'text' ? message.content[0].text : ''
}

// Usage
const brief = await generateContentBrief('Bitcoin liquidation calculator')
console.log(brief)

async function generateMetaTags(
  title: string,
  content: string
): Promise<{ metaTitle: string; metaDescription: string }> {

  const message = await client.messages.create({
    model: 'claude-3-haiku-20240307', // Use Haiku for cost efficiency on bulk tasks
    max_tokens: 200,
    messages: [{
      role: 'user',
      content: `Based on this blog post title and content summary, write:
1. An SEO meta title (under 60 characters, lead with primary keyword)
2. An SEO meta description (155-160 characters, include the keyword, end with a benefit or action)

Title: ${title}
Content summary: ${content.substring(0, 500)}

Respond in JSON format: { "metaTitle": "...", "metaDescription": "..." }`
    }]
  })

  const text = message.content[0].type === 'text' ? message.content[0].text : '{}'
  return JSON.parse(text)
}

async function generateFAQWithSchema(
  topic: string,
  postContent: string
): Promise<{ faqHtml: string; jsonLd: object }> {

  const message = await client.messages.create({
    model: 'claude-3-5-sonnet-20241022',
    max_tokens: 3000,
    messages: [{
      role: 'user',
      content: `Generate 8 FAQ questions and answers for a blog post about: ${topic}

Context: ${postContent.substring(0, 1000)}

Requirements:
- Questions should match "People Also Ask" patterns
- Answers should be 50-100 words each (optimal for Featured Snippets)
- Include specific numbers or facts where possible

Return JSON with this exact structure:
{
  "faqs": [
    { "question": "...", "answer": "..." },
    ...
  ]
}`
    }]
  })

  const text = message.content[0].type === 'text' ? message.content[0].text : '{"faqs":[]}'
  const { faqs } = JSON.parse(text)

  // Generate JSON-LD
  const jsonLd = {
    '@context': 'https://schema.org',
    '@type': 'FAQPage',
    mainEntity: faqs.map((faq: { question: string; answer: string }) => ({
      '@type': 'Question',
      name: faq.question,
      acceptedAnswer: { '@type': 'Answer', text: faq.answer }
    }))
  }

  // Generate HTML
  const faqHtml = faqs.map((faq: { question: string; answer: string }) =>
    `<h3>${faq.question}</h3><p>${faq.answer}</p>`
  ).join('')

  return { faqHtml, jsonLd }
}

// Always guard local development
const IS_DEV = process.env.NODE_ENV === 'development'
if (IS_DEV) return DUMMY_RESPONSE