菜单
Use when analyzing one webpage's primary topic, finding DataForSEO-backed keyword opportunities, or producing evidence-based search, Open Graph, and Twitter metadata in a Markdown report.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Use when creating or maintaining a persistent SEO keyword watchlist, discovering ranked domain keywords, checking Google positions across runs, comparing gains and losses, listing or removing tracked entries, or producing a ranking-change report.
Use when consolidating dated SEO reports for one domain into a concise full SEO report, prioritizing critical next actions, or surfacing cross-analysis ranking opportunities from files under SEO/<domain>/ with a current DataForSEO MCP validation snapshot.
Use when auditing one webpage for technical SEO, indexability, links, metadata, structured data, or desktop/mobile rendering and performance, then producing a prioritized report with DataForSEO-derived Technical Scores.
Use when performing SEO analysis with the DataForSEO MCP, selecting or sequencing DataForSEO tools, interpreting their responses, estimating request scope, troubleshooting calls, or explaining the MCP and underlying API contracts.
Use when detecting pages competing for the same keywords, selecting canonical target pages, deciding whether to merge, redirect, canonicalize, differentiate, or retain overlapping content, or recommending evidence-based internal links with DataForSEO MCP.
Use when detecting declining organic content, separating ranking decay from seasonality, prioritizing existing pages for refresh, or producing evidence-based content refresh briefs with DataForSEO MCP.
| name | seo-page-metadata |
| description | Use when analyzing one webpage's primary topic, finding DataForSEO-backed keyword opportunities, or producing evidence-based search, Open Graph, and Twitter metadata in a Markdown report. |
| compatibility | Requires filesystem write access and the official DataForSEO MCP server with ONPAGE and DATAFORSEO_LABS enabled. Python 3 is recommended for deterministic validation and normalization. |
Analyze one supplied page without treating its existing metadata or URL as proof of its topic. Use DataForSEO MCP for every SEO data call.
python3 <this-skill-directory>/scripts/metadata_support.py filename --url <url> --date <local-date> to validate it and obtain the normalized domain and filename. Otherwise apply the equivalent input and path contract directly; do not stop solely because Python is unavailable. Ask and wait when the URL is absent or invalid.on_page_instant_pages, on_page_content_parsing, and dataforseo_labs_google_related_keywords; active schemas override captured examples. Announce the normal seven-call budget: two OnPage calls and five Related Keywords calls, each with depth: 2 and limit: 25.structured_main_topic or projection_degraded_text, using the helper when available or the documented response paths and filters directly. Derive topic, intent, geography, seeds, and suggestions only from qualifying Content Parsing evidence.projection_degraded_text, disclose the lower confidence before keyword calls and stop instead when the fallback text is too noisy or thin to justify all five seeds. Ask before a JavaScript retry or other billable recovery.scripts/metadata_support.py when Python 3 is available; otherwise perform the documented normalization, merge, validation, and ranking rules directly. Preserve null versus zero, merged seed provenance, metric conflicts, exclusions, coverage, and the derived opportunity_proxy formula in either mode.<cwd>/SEO, return its absolute path, and summarize the recommendation, keyword coverage, limitations, and total cost.Ask before every retry, pagination request, extra call, or more expensive JavaScript rendering. Never bypass MCP, invent missing metrics, or persist raw customer responses in the repository.