build_segment
ChatGPTDeterministically validates a structured segment definition and returns an opaque, store/user-scoped token for compatible Metorik data tools. Call list_store_metadata with type=segment_options for the same resource first. This tool does not make an AI generation call.
get_engage_campaigns
ChatGPTLists Engage campaigns or returns one campaign with optional stats. Use this for Engage automations, broadcasts, cart automations, campaign status, send performance, and campaign-level stats. Use include_content=true with campaign_id to retrieve the current saved template with unresolved placeholders; content is omitted by default.
get_goals
ChatGPTLists or retrieves store-wide goals with current target progress, achieved/missed/in-progress status, and optional period snapshots.
get_store_catalog
ChatGPTReturns sales-oriented catalog lists for products, variations, categories, brands, product groups, or coupons over a date range. Use this for best-selling products, rankings, comparisons, and named catalog lookup. For each variation of a product or a breakdown by its color, size, finish, style, or option, first find the parent in products, then query variations with product_id set to the returned parent ID and per_page=100; this applies an exact parent filter and avoids a store-wide variation list. Continue identical pages while has_more_pages is true before calculating complete attribute totals. Present a complete one-page result as one ranked list, without a chart or extra actions unless the user asks for them. For one named entity over time, pass its public identifier to get_store_catalog_by_date. Variation revenue must come from variation rows such as net_sales, never whole-order revenue from orders that happen to contain a matching variation.
get_store_catalog_by_date
ChatGPTReturns time-series sales data for one resolved product, variation, category, brand, product group, or coupon over a date range. Use this after identifying a catalog item; for a named entity, first search the plural resource with get_store_catalog and pass the selected row's public identifier here. Use gross_items_sold for how many, units, or items sold; use order count only when explicitly requested.
get_store_report
ChatGPTAnswers questions such as which sizes sell best, red shirt sales, and popular colour/size combinations: discover stored names with list_store_metadata type=variation_attributes first. Returns consolidated analytics for variation attribute sales (report_type=variation_attributes), time series, grouped, source, breakdown, subscription plan, cart product ranking, customer cohort, custom metric, or products-bought-together reports. Use this as the default analytics/reporting tool before record-level tools for store-wide sales/revenue over time or other order, customer, subscription, and cart analytics. Products-bought-together rows include times_bought_together and total_order_count for product_1 in the same period and filters. For most abandoned products, use report_type=carts, report=products, and order_by=abandoned_count. A daily, weekly, monthly, or yearly trend across a date range requires one timeseries call with the matching group_by, never one call per time bucket. For Net Sales use the orders time-series report; for Net Revenue use the revenue time-series report. Do not use it for one named product, variation, category, brand, product group, or coupon over time; resolve that entity with get_store_catalog and use get_store_catalog_by_date.
get_store_totals
ChatGPTReturns one aggregate count/total result for orders, customers, subscriptions, refunds, or carts after optional search, filters, or a saved segment. Use this when the user asks how many records match rules or wants aggregate totals without individual records. For customers, omit order_start_date and order_end_date for lifetime statistics, or provide both to calculate, filter, and total customer order statistics within that period. Segmented customer totals include current active, churned, classified, and churn-share subscription values when subscription data is available. Do not use it repeatedly to build daily, weekly, or monthly time series; use get_store_report with report_type=timeseries and group_by instead.
list_store_metadata
ChatGPTLists supporting metadata such as variation attribute names (type=variation_attributes), report options, segment options, saved segments, custom metrics, object fields, Metorik Property keys (type=properties, resource=customers), and top-level platform resource meta keys. Use segment_options with a concise search term before authoring a structured definition for build_segment.
list_store_objects
ChatGPTReturns paginated order, customer, subscription, refund, or cart records with optional search, filters, fields, and sorting. Use this only when the user needs individual records, not aggregate counts or report totals.
list_stores
ChatGPTLists stores the authenticated user can access, including the current default store and whether record-level MCP object data is enabled. Use this first for OAuth connections unless the user already provided a store_id.
search_store
ChatGPTSearches store resources like customers, orders, subscriptions, products, variations, categories, coupons, and refunds. Use this to find specific known records by text, such as a customer name, email, order number, product name, or coupon code. Synced notes are unavailable in this tool; use list_store_objects with include_notes=true, which also supports search and filters.
build_segment
Claudeget_engage_campaigns
Claudeget_goals
Claudeget_store_catalog
Claudeget_store_catalog_by_date
Claudeget_store_report
Claudeget_store_totals
Claudelist_store_metadata
Claudelist_store_objects
Claudelist_stores
Claudesearch_store
Claude
