rankm8-project-setup
Use this skill when a user wants to create a new RankM8 SEO project and run the complete headless setup pipeline through the RankM8 MCP server.
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。
菜单
Use this skill when a user wants to create a new RankM8 SEO project and run the complete headless setup pipeline through the RankM8 MCP server.
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 this skill when a user wants to create SEO analysis plans, outlines, full content, meta information, or AI images with RankM8 MCP generation tools.
Use this skill when a user wants to discover, validate, import, analyze, link, rank, or delete SEO keywords through RankM8 MCP.
Use this skill when a user wants a monthly SEO performance report from RankM8 projects, Search Console, Google Business Profile, and tracked keywords.
Use this skill when a user wants to run, inspect, triage, or update technical SEO audits for RankM8 project pages via MCP.
| name | rankm8-project-setup |
| description | Use this skill when a user wants to create a new RankM8 SEO project and run the complete headless setup pipeline through the RankM8 MCP server. |
Guide an AI client through a complete RankM8 project setup without opening the web UI. The playbook creates the project, discovers pages, crawls selected pages, enriches keywords, runs keyword analysis, and stores goals plus competitors.
Use the MCP server as the source of truth. Keep the project_id returned by
create_project and pass it to every later project-scoped call.
Setup tools:
analyze_website(url)create_project(name, project_url, company_name, company_location, company_country, language, project_scope, coordinates?)scrape_sitemap(project_id, sitemap_url?)serp_site_search(project_id)crawl_pages(project_id, sitemap_ids)fetch_keyword_data(project_id)suggest_selection(project_id, type, limit?, sitemap_ids?)list_competitor_candidates(project_id, keyword_ids?, max_position?, limit?)get_setup_status(project_id)update_project(project_id, name?, company_name?, company_location?, company_country?, language?, project_scope?, coordinates?, company_description?, business_goals?, selected_competitors?, selected_sitemap_ids?, selected_keyword_ids?)Pipeline handoff tool:
analyze_keywords(project_id, keyword_ids)analyze_website(url) to infer company_name,
company_location, company_country, and language.create_project(name, project_url, company_name, company_location, company_country, language, project_scope, coordinates?).project_id for every later call.scrape_sitemap(project_id, sitemap_url?).get_setup_status(project_id) every 15-30 seconds until
steps.sitemap_scrape.status is completed or completed_with_failures.steps.sitemap_scrape.status is failed, if pages_found is 0, or if
the response suggests a fallback, ask the user before the paid fallback and
call serp_site_search(project_id).suggest_selection(project_id, type="pages", limit=10).sitemap_ids list, ideally 5-10 pages and never more than 25.crawl_pages(project_id, sitemap_ids=[...]).get_setup_status(project_id) every 15-30 seconds until
steps.keyword_extraction.status is completed or completed_with_failures.fetch_keyword_data(project_id).get_setup_status(project_id) every 15-30 seconds until
steps.keyword_data.status is completed or completed_with_failures.suggest_selection(project_id, type="keywords", limit=10, sitemap_ids=[...]) when page-scoped suggestions are useful.keyword_ids.update_project(project_id, selected_keyword_ids=[...]) before keyword
analysis so setup status is scoped to the confirmed keywords.analyze_keywords(project_id, keyword_ids=[...]).get_setup_status(project_id) every 15-30 seconds until
steps.keyword_analysis.status is completed or completed_with_failures.list_competitor_candidates(project_id, keyword_ids=[...], max_position=10, limit=10).more_traffic, more_leads,
higher_sales, brand_awareness, local_visibility.update_project(project_id, business_goals=[...], selected_competitors=[...], selected_sitemap_ids=[...], selected_keyword_ids=[...]).Treat get_setup_status(project_id) as the only setup polling tool. Watch
overall_status, current_step, steps, poll_again_in_seconds, next_action,
and warning.
Valid step statuses are not_started, pending, running, completed,
completed_with_failures, and failed.
Never poll forever. If a response contains warning, explain it to the user and
either retry the indicated step tool, switch to a documented fallback, or stop.
If any step is failed, stop the pipeline and handle the error before continuing.
Always get explicit user confirmation before:
analyze_website, scrape_sitemap,
serp_site_search, crawl_pages, fetch_keyword_data, analyze_keywords.create_project with final project data.sitemap_ids.keyword_ids.selected_competitors and business_goals.GSC, GMB, and WordPress connections cannot be completed through MCP. Refer the user to project settings in the RankM8 web UI for those browser OAuth flows.
project_limit_reached: Tell the user their plan limit is reached and suggest a
plan upgrade or deleting unused projects.insufficient_credits: Stop before the paid step and ask the user to add credits.invalid_parameter: Show the invalid field and request corrected input.invalid_url or missing_url: Ask for a public HTTP(S) website URL.invalid_sitemap_ids or invalid_keyword_ids: Re-list pages or keywords and ask
the user to choose IDs from the current project.missing_sitemap_ids or too_many_pages: Ask for 1-25 page IDs.analysis_failed, search_failed, or dispatch_failed: Stop, summarize the
failed step, and retry only after the user approves.project_not_found: Re-run project selection or ask the user to reconnect OAuth.