将 Semrush MCP 连接到 AI
全球最强大的流量、可见度和市场数据集。已融入您的 AI 助手中。
立即获取门票
为期一天。实战策略。专为志在必得的营销人打造。
2026 年 10 月 13 日英国伦敦

Workflow: Create briefs and drafts for refreshes

Pull the current page content and its SEO performance baseline before making any changes. This step documents what exists so the refresh has a clear starting point.

Add your data

Adding details ensures a relevant result
Save time by automatically applying details and getting ready-to-use prompts.

Prompt

For each page selected for content refresh, extract the current content and SEO performance baseline for {your-domain.com} in {country}. Pages to analyze: * If {paste URLs here} are provided, include those pages. * If a refresh backlog already exists in this conversation, also include all pages marked `full_rewrite` or `targeted_update`. * If both sources are available, combine them and remove duplicate URLs. * If neither {paste URLs here} nor a refresh backlog is available, ask the user to confirm which page URLs they want to refresh before proceeding. For each page: 1. Fetch the current page content. 2. Pull the page's current keywords, positions, and estimated traffic ({country}). Return ONE table: Columns: * page_url * source (provided_url / refresh_backlog / both) * word_count * h1 * meta_description (first 160 chars) * top_3_keywords (keyword: position) * estimated_monthly_traffic * content_issues (thin / outdated / missing_keywords / none) Content issue rules: * thin = word_count < 800 * outdated = content references specific years prior to {current_year} - 2 * missing_keywords = page ranks for fewer than 5 keywords * none = no obvious issues detected If multiple issues apply, list all separated by commas. Complete all calls before returning the final table.

Example Output (illustrative)

- A table per page showing: word count, H1, meta description excerpt, top 3 ranking keywords with positions, estimated traffic, and content issue flags (thin / outdated / missing_keywords)