SEO MCPSEO MCP Tools Reference
Find SEO MCP tools for projects, crawls, pages, links, images, Rank Tracker, AI Visibility, uptime, and Core Web Vitals.
SEO MCP provides focused tools for reading existing Screpy data and requesting supported account changes. Tool availability can change with the product and account access; use the connected client's tool list as the current inventory. The server uses discoverable public identifiers, validated filters, and permission checks.
Tool inventory
| Category | Tools |
|---|---|
| Account | get_account_context, get_plan_usage |
| Projects | list_projects, get_project_overview, compare_projects, create_project, update_project_settings, delete_project |
| Crawls | list_crawls, get_crawl_status, get_crawl_summary, get_on_page_overview, compare_crawls, start_crawl |
| Pages | list_pages, get_page_details, list_quick_wins |
| Links | list_links, get_link_sources |
| Images | list_images, get_image_sources |
| Monitoring | get_uptime_summary, list_core_web_vitals |
| Rank Tracker | get_rank_tracker_options, get_rank_tracker_overview, list_rank_tracker_keywords, get_rank_tracker_keyword, list_rank_tracker_segments, preview_rank_tracker_segment, create_rank_tracker_segment, update_rank_tracker_segment, delete_rank_tracker_segment, add_rank_tracker_keywords, delete_rank_tracker_keywords |
| AI Visibility | get_ai_visibility_options, get_ai_visibility_overview, list_ai_visibility_prompts, get_ai_visibility_prompt_activity, list_ai_visibility_mentions, list_ai_visibility_sources, get_ai_visibility_source_activity, get_ai_visibility_answer, add_ai_visibility_prompts, delete_ai_visibility_prompt |
Required discovery order
Use list_projects before any project tool. Use list_crawls before tools that require a crawl UID. Rank Tracker and AI Visibility use their own analysis records; they do not require a crawl UID. Never guess an identifier.
list_projects
→ list_crawls(project_uid)
→ get_on_page_overview(project_uid, crawl_uid)
→ list_pages(project_uid, crawl_uid, filters)This order keeps tool calls reproducible and prevents accidental use of internal or stale identifiers.
Pagination and filters
Large record sets use bounded results and, where exposed by a tool, cursor pagination. Crawl pages, links, and images accept structured filters. Supported operators include:
eq,neqgt,gte,lt,lteincontainsis_null
Use no more than 20 filters in one request. Keep each query focused enough that the returned records answer one clear question.
Read-only and state-changing tools
The analysis surface is read-only. These tools change state:
create_projectupdate_project_settingsdelete_projectstart_crawladd_rank_tracker_keywords,delete_rank_tracker_keywordscreate_rank_tracker_segment,update_rank_tracker_segment,delete_rank_tracker_segmentadd_ai_visibility_prompts,delete_ai_visibility_prompt
The server instructs agents to avoid them unless the user explicitly requests the action. High-impact operations require explicit confirmation.
For field-level examples, open Rank Tracker, AI Visibility, or another category in the sidebar. For a task-oriented starting point, use the workflow hub.
Related documentation