77 tools your agent can call.

Which tools a given account sees depends on the plan, the Google permissions granted and the keys added (Bing, IndexNow). Write tools are marked write: they need write mode on a paid plan and, for Google, a separate consent. Everything else is read-only and safe to let an agent call freely.

Google Search Console

Performance data, URL inspection and sitemaps through the Search Console API, with OAuth handled for you.

• gsc_list_sites List every Search Console property the account can access, with permission level.
• gsc_get_site Get one property and the account's permission level on it.
• gsc_query_analytics Search performance grouped by query, page, country, device, search appearance, date or hour, with filters. Up to 25,000 rows per query.
• gsc_list_sitemaps Sitemaps submitted for a property, with status, errors, warnings and indexed counts.
• gsc_get_sitemap Details of one submitted sitemap.
• gsc_inspect_url URL Inspection: indexed or not, coverage state, canonicals, last crawl, robots, mobile usability and rich results.
• gsc_inspect_urls_batch Inspect up to 500 URLs in one call, throttled to stay inside Google's quota, with a compact status per URL.
• gsc_add_sitewrite Add a property to Search Console (verification still happens in the Search Console UI).
• gsc_delete_sitewrite Remove a property from the account's Search Console.
• gsc_submit_sitemapwrite Submit or resubmit a sitemap.
• gsc_delete_sitemapwrite Remove a sitemap from Search Console.
• gsc_request_indexingwriteoptional Notify Google about an updated or deleted URL through the Indexing API (officially for JobPosting and BroadcastEvent pages).

SEO reports

Analyses that run on our servers and return finished numbers, so the agent reasons about results instead of paging through raw rows.

• report_quick_wins Query and page pairs ranking in positions 4–20 with many impressions: the cheapest traffic to gain, sorted by potential extra clicks.
• report_traffic_drops Compare a period with the same-length period before it and return the biggest click losers and gainers by page or query.
• report_content_decay Pages whose clicks fell in each of the last three windows: a steady slide, not a one-off dip. Candidates for a content refresh.
• report_cannibalization Queries where two or more of your pages each get a meaningful share of impressions, splitting ranking signals.
• report_ctr_outliers Pages whose CTR is under half of your site's median at the same position: candidates for better titles and descriptions.
• report_brand_split Branded vs non-branded queries with clicks, impressions and click share, plus the top queries in each group.
• report_index_health Read URLs from your sitemaps, inspect them with the URL Inspection API and group them by coverage state.
• report_google_vs_bing Clicks and impressions per page from Google and Bing side by side.

Bing Webmaster Tools

The full Bing Webmaster API: search performance, crawl data, backlinks, keyword research and URL submission.

• bing_list_sites Sites in Bing Webmaster Tools for the API key, with verification status.
• bing_get_site_roles Users with access to a site.
• bing_query_stats Clicks, impressions and average position for top search queries (weekly, about six months).
• bing_page_stats Clicks, impressions and average position for top pages.
• bing_page_query_stats Search queries that led to one page.
• bing_query_page_stats Pages shown for one search query.
• bing_query_page_detail_stats Detailed stats for one query on one page.
• bing_query_traffic_stats Traffic history for one query.
• bing_rank_traffic_stats Daily clicks and impressions for the whole site.
• bing_crawl_stats Daily crawled pages, crawl errors, in-index counts and pages blocked by robots.txt.
• bing_crawl_issues URLs with crawl issues: 4xx/5xx, DNS, timeouts, malware.
• bing_url_info Index details for a single URL: HTTP status, last crawl, discovery date, size, anchors.
• bing_url_traffic Clicks and impressions for a single URL.
• bing_list_sitemaps Sitemaps and feeds submitted to Bing, with status and URL counts.
• bing_sitemap_details Child sitemaps of a sitemap index.
• bing_link_counts Your pages that have inbound links, with link counts.
• bing_url_links External pages linking to a URL on your site.
• bing_keyword Bing search impressions for any keyword in a period (keyword research, not limited to your site).
• bing_keyword_stats Historical weekly impressions for a keyword.
• bing_related_keywords Related keywords with impressions for a period.
• bing_url_submission_quota Remaining daily and monthly URL submission quota.
• bing_list_blocked_urls URLs and directories blocked from Bing results.
• bing_crawl_settings Crawl rate per hour and crawl boost setting.
• bing_list_query_params Query parameters Bing ignores for URL normalization.
• bing_fetched_urls URLs previously fetched with Fetch as Bingbot.
• bing_fetched_url_details Result of a Fetch as Bingbot request: status, headers, content.
• bing_add_sitewrite Add a site to Bing Webmaster Tools.
• bing_verify_sitewrite Attempt verification of a site (meta tag, XML file or DNS already in place).
• bing_remove_sitewrite Remove a site from Bing Webmaster Tools.
• bing_submit_urlswrite Submit up to 500 URLs for crawling and indexing, checking the remaining quota first.
• bing_save_crawl_settingswrite Set the hourly crawl rate and crawl boost.
• bing_submit_sitemapwrite Submit a sitemap or feed to Bing.
• bing_remove_sitemapwrite Remove a submitted sitemap or feed.
• bing_block_urlwrite Block a page or directory from Bing results.
• bing_unblock_urlwrite Remove a page or directory block.
• bing_add_query_paramwrite Tell Bing to ignore a query parameter.
• bing_remove_query_paramwrite Stop ignoring a query parameter.
• bing_fetch_urlwrite Ask Bingbot to fetch a URL now; read the result later with bing_fetched_url_details.

IndexNow

Tell Bing, Yandex, Seznam, Naver and other IndexNow engines about new, changed or deleted URLs within seconds.

• indexnow_generate_key Generate a new IndexNow key and the file you need to host on the site.
• indexnow_submitwrite Notify IndexNow engines about new, changed or deleted URLs: one host per call, up to 10,000 URLs.

Core Web Vitals and PageSpeed

Real-user Chrome UX Report data and Lighthouse audits for any origin or URL you own.

• cwv_get_origin Real-user Core Web Vitals (LCP, INP, CLS, FCP, TTFB) for a whole origin, last 28 days, p75 and good/needs-improvement/poor shares.
• cwv_get_url Real-user Core Web Vitals for a single URL (needs enough Chrome traffic).
• cwv_history Weekly Core Web Vitals history, up to about six months, for an origin or URL.
• psi_run_audit Run Lighthouse through PageSpeed Insights: category scores, lab metrics, field data and top opportunities.

Google Analytics 4

GA4 reports next to your search data: sessions, key events, landing pages and referral traffic from AI assistants.

• ga_list_properties Google Analytics accounts and GA4 properties the account can access.
• ga_list_data_streams Data streams of a GA4 property, including the Measurement ID used by the tag on the site.
• ga_get_metadata Dimensions and metrics available for a property, including custom ones, with search.
• ga_run_report Run a GA4 report with dimensions, metrics, date ranges, filters, ordering and paging.
• ga_run_realtime_report GA4 realtime report for the last 30 minutes.
• ga_list_key_events Key events (formerly conversions) configured on a property.

Google Tag Manager

Read-only view of what is actually firing on the site: containers, tags, triggers, variables and the live version.

• gtm_list_accounts Tag Manager accounts the account can access.
• gtm_list_containers Containers of an account, with the public GTM-XXXX id.
• gtm_list_workspaces Workspaces (drafts) of a container.
• gtm_list_tags Tags of a workspace with type, triggers and parameters, including GA4 measurement ids and event parameters.
• gtm_list_triggers Triggers of a workspace with type and conditions.
• gtm_list_variables User-defined variables of a workspace.
• gtm_get_live_version The currently published container version: what is really running on the site.

How the tools are designed

• Finished numbers, not raw dumps. Reports compare periods, compute shares and sort by impact on our servers, so the model reads a short result instead of 25,000 rows.
• Explicit limits. Results carry truncation flags and quota notes (URL Inspection: 2,000 a day) so the agent never treats a partial dataset as complete.
• Portable schemas. No unions, references or free-form records in input schemas, so every tool loads in Claude, ChatGPT, Cursor and smaller clients alike.
• Safety hints. Each tool declares readOnlyHint, destructiveHint and openWorldHint, and the server enforces read-only mode regardless of what the client asks.

Connection details for developers are on the any MCP client page. The endpoint is https://mcp.gscmagic.com/mcp.