MCP App Store
by KeenEthics
Clarmix icon

Clarmix MCP Server for Claude and ChatGPT

By Clarmix

Overview

Clarmix connects Claude to your Google Ads, Google Merchant Center (GMC) and Shopify data so you can analyze and improve your e-commerce advertising in plain language. Built for PPC managers, agencies and online merchants running Google Shopping and Performance Max (PMax) campaigns. What you can do: • Google Ads: account and campaign performance, ROAS vs target (tROAS), budget-limited campaigns, search terms, Shopping and PMax structure, conversion tracking setup, change history and custom GAQL queries. • Product performance: find which products are Profitable, Costly, Zero Conversion or Zombies, follow label trends over time, and check that Clarmix custom labels are actually used in your Shopping and PMax campaigns. • Google Merchant Center: account issues, product disapprovals, data sources and product feed health. • Product feed quality: content and GEO scores, weak titles, descriptions and attributes, and AI enrichment setup for your product data. • Shopify: sales, inventory, profitability and margins (POAS), plus catalog gaps between Shopify, Merchant Center and Google Ads. Example prompts: "Which products wasted the most ad spend last month?" "Why are my products disapproved in Merchant Center?" "Are my Clarmix labels used in my PMax asset groups?" "Which Shopify best-sellers get no Google Shopping impressions?" Clarmix is read-only. Claude can analyze your data but never changes campaigns, budgets, feeds or products. You make changes in Clarmix, Google Ads or Merchant Center. Requires a Clarmix account (app.clarmix.com) with Google Ads and/or Merchant Center connected. Shopify is optional.

How to connect

  1. Claude: Open the Clarmix listing in the Claude connectors directory and choose Connect. Claude listing Clarmix exposes 27 tools here.
  2. ChatGPT: Open the Clarmix listing in the ChatGPT app directory and choose Connect. ChatGPT listing Clarmix exposes 35 tools here.

Tools

ads_account_performance

ChatGPT
Live Google Ads totals (cost, clicks, impressions, conversions, conversion value, ROAS, CPA, CTR, CPC) with a daily/weekly/monthly time series, optionally broken down by campaign, campaign type, network or device (totals per value plus a series per value, e.g. weekly cost per campaign type). Use campaignIds to get the time series of specific campaigns (e.g. before/after a change).

ads_change_history

ChatGPT
Who changed what in Google Ads, and when. Last 30 days: every change with user email, tool (web UI, Editor, API, scripts, automated rules, recommendations), campaign / ad group, and decoded before → after values (target ROAS, target CPA, budget, status, name, bids, bidding strategy…), plus counts by user, campaign and field. Up to 90 days: activity (number of resources changed per campaign, resource type and day, including PMax asset group listing groups / assets and changes made by Google, which the detailed history does not contain; no user or values). Older than 90 days: Google Ads UI only.

ads_conversion_setup

ChatGPT
Conversion actions (status, category, primary/secondary, counting, volume), account-level goals, and the biddable goal sets used by enabled campaigns (campaigns grouped by identical goals).

ads_list_campaigns

ChatGPT
Google Ads campaigns (Search, Shopping, Performance Max, Demand Gen, Display, Video…) with status, bidding strategy, target ROAS/CPA, budget, feed label, metrics and search impression share (lost to budget / rank) for the period, sorted by cost. By default only active campaigns (enabled, or with impressions in the period); hidden counts the others.

ads_pmax_product_mapping

ChatGPT
Which products are in which Performance Max asset group. Takes the products in scope of each PMax campaign (Google Ads shopping_product, after the campaign's Merchant Center / feed filters) and evaluates each enabled asset group's listing-group tree on them. Returns per asset group the product count (eligible, split by Clarmix label, sample item ids), products in a campaign's scope that no asset group includes (unassigned), products in several asset groups of the same campaign, and a paged product list (item id, title, status, asset groups). Enabled PMax campaigns only unless campaignIds is given.

ads_pmax_structure

ChatGPT
Performance Max asset groups with ad strength, metrics and what each one targets (listing-group filters: custom labels, brands, product types, item ids), sorted by cost. Also shows how Clarmix labels (clarmix_* values on the account custom label slot) are split across enabled asset groups, which labels are not targeted, and asset groups targeting all products. Enabled groups of enabled campaigns only by default.

ads_product_performance

ChatGPT
Live product-level Google Ads performance from Shopping and Performance Max (shopping_performance_view) for any date range: cost, revenue, ROAS, conversions, CPA, clicks, impressions, CTR, CPC, conversion rate. Group by one or several dimensions (e.g. ["product", "country"] for each product per country): product, campaign, country, language, channel (online / local), device, brand, product type L1-L5, Google category L1-L3, custom labels. Filter by offer ids, product title text, brands, product type text, countries (ISO codes), languages, channel, Clarmix labels, custom label values or campaigns; apply metric thresholds (e.g. ROAS > 4, cost > 50) and sort. granularity adds a day / week / month series per row (up to 20 rows; filter to a few products first).

ads_search_terms

ChatGPT
Search terms that triggered ads: Search/Shopping (search_term_view, with ad group and added/excluded status) and Performance Max search term insights (campaign_search_term_view). Filter by campaign, text, and metric thresholds. Up to 1000 terms per source (highest cost first) are fetched; page through them with limit + offset (see nextOffset) to build full lists.

ads_shopping_structure

ChatGPT
Standard Shopping ad groups with metrics and what each one targets (product partitions by custom label, brand, product type, item id, with bids and exclusions in detail "full"), sorted by cost, plus how Clarmix labels are used in the partitions. Enabled ad groups of enabled campaigns only by default.

check_label_delivery

ChatGPT
Checks the whole Clarmix label chain for one account, step by step: (1) labels computed in Clarmix, (2) label feed published by Clarmix, (3) feed connected to each Merchant Center feed, (4) labels applied in Merchant Center (current Clarmix label vs the value in the custom label slot), (5) labels used in Google Ads (PMax asset groups / Shopping product groups that split on the Clarmix labels, and their share of spend). Each step has a status, a finding in plain words and where to fix it (Clarmix, Merchant Center or Google Ads). Answers "are my Clarmix labels working / used in my campaigns?".

ga_keywords

ChatGPT
mode "paid": Google Ads keywords or search queries as seen in GA4 (needs the GA–Google Ads link), with sessions, engagement rate, purchases and revenue. mode "site_search": terms shoppers type in the website search (view_search_results events), with search counts and users. Organic Google queries are not available in GA4. Returns a "not connected" message when Google Analytics is not linked.

ga_landing_pages

ChatGPT
GA4 landing pages (path without query string) with sessions, engagement rate, key events, add-to-carts, purchases, revenue and purchase rate, for all traffic, Google Ads paid traffic (google / cpc) or organic search. Returns a "not connected" message when Google Analytics is not linked.

ga_product_funnel

ChatGPT
Per product (GA item id): item views, add-to-carts, purchases, item revenue, view→cart and cart→purchase rates, joined with the Clarmix product (title, Clarmix label) by Merchant Center offer id and, when Google Ads is connected, its Shopping / PMax clicks, cost and conversions over the same period. Flags products with many ad clicks but few add-to-carts and products viewed but never bought. Items that do not match a Clarmix product are counted apart. Returns a "not connected" message when Google Analytics is not linked.

ga_property_overview

ChatGPT
GA4 property linked to the Clarmix account and its ecommerce tracking health: currency, time zone, website vs the Merchant Center website, whether the property is linked to the account's Google Ads and Merchant Center, key events configured, the last 7 complete days of view_item / add_to_cart / begin_checkout / purchase events, purchases and revenue, and the share of item views and purchases tracked without an item id. Returns a "not connected" message with setup steps when Google Analytics is not linked.

ga_query

ChatGPT
Runs a custom GA4 Data API report (runReport) on the account's linked property: dimensions and metrics by GA4 API name (custom dimensions allowed), optional dimensionFilter / metricFilter in GA4 FilterExpression JSON, orderBys, limit (max 1000) and offset. Identifier dimensions (clientId, user-scoped custom dimensions, transactionId) are not available. Returns rows, rowCount, metadata (thresholding, currency, time zone), quota, or the GA error so the query can be fixed. Returns a "not connected" message when Google Analytics is not linked.

ga_tracking_health

ChatGPT
Compares, per day, GA4 purchases and revenue with Google Ads conversions and conversion value (all campaigns, and Shopping + Performance Max only) and, when Shopify is connected, Shopify orders and net sales. Returns gap percentages, days where GA has no purchase while Ads or Shopify do, duplicated transaction counts (no ids), purchases without transaction id, and the share of GA item ids that match a Clarmix / Merchant Center offer id (with up to 20 unmatched examples). Returns a "not connected" message when Google Analytics is not linked.

ga_traffic_acquisition

ChatGPT
GA4 sessions, engaged sessions, engagement rate, key events, purchases, revenue and purchase rate grouped by default channel group, source / medium, campaign, or Google Ads campaign (needs the GA–Ads link), with an optional comparison to the previous period. Returns a "not connected" message when Google Analytics is not linked.

get_account_overview

ChatGPT
Clarmix settings for one account (ROAS / cost-per-conversion targets, default lookback window, conversion lag, which custom_label slot carries Clarmix labels, credits) plus the latest cached Google Ads and Merchant Center summaries (global metrics, label distribution, content quality averages), and which data sources are connected to Clarmix (Google Ads, Merchant Center, Shopify, data files) with what each missing one would unlock.

get_content_quality

ChatGPT
Clarmix content-quality scores for the catalog (title, description, Google category, product type, highlights, details, Q&A, item group title, related products, document links, variant options, popularity rank): averages, tier distribution, custom label usage, the weakest products for a chosen attribute, and an enrichment plan: weak attributes ranked by ad spend on poor-content products, with the Clarmix action and credit cost for each.

get_enrichment_setup

ChatGPT
How AI enrichment is configured for this account: custom prompt rules per attribute (the instructions Clarmix gives the AI), recent enrichment tasks, available credits, imported data sources (CSV/XML) and the catalog of external/Shopify properties available for rules.

get_labels_and_rules

ChatGPT
How Clarmix labels products for this account: label definitions (conditions, priority, wasted/profitable spend), custom rules that override ROAS / cost-per-conversion targets for product segments, the evaluation order, and how labels are pushed to Merchant Center (custom_label slot + value format) so they can be used in Google Ads listing groups. labelDelivery summarizes whether this account's labels really reach Merchant Center (label feed linked, labels applied, or labels coming from another Clarmix account), i.e. whether these label rules are the ones applied to the Google Ads products.

get_product

ChatGPT
Everything Clarmix knows about one product: Merchant Center attributes, original vs AI-enriched vs manual values, all content scores with reasons, external data (Shopify cost/margin/stock/orders, CSV columns), stored Ads metrics and label history.

gmc_account_issues

ChatGPT
Live account-level Merchant Center issues and alerts (suspensions, policy violations, missing settings) with severity, affected destinations/countries and documentation links.

gmc_data_sources

ChatGPT
Primary and supplemental data sources (feeds) in Merchant Center, including the Clarmix supplemental label/enrichment feeds, with feed label, language, countries and fetch schedule.

gmc_product_issues

ChatGPT
Live product health in Merchant Center: active / pending / disapproved / expiring counts per destination and country, the most frequent item-level issues (with resolution hints and number of affected products), and optionally sample products affected by an issue.

gmc_product_performance

ChatGPT
Clicks, impressions and CTR per product from Merchant Center (free listings and ads, by marketing method), for a date range.

list_accounts

ChatGPT
Lists every Clarmix account the user can access (one account = one Google Ads account + its Merchant Center). Each account has a Clarmix "accountId", the account identifier used across Clarmix. "you" is the connected Clarmix user.

run_gaql

ChatGPT
Runs any Google Ads Query Language SELECT on this account (ad groups, keywords, assets, audiences, geo, devices, bidding strategies…). Single SELECT only, max 1000 rows, the customer is fixed by accountId. Money fields ending in _micros are returned converted (cost_micros → cost). Always add a segments.date filter when selecting metrics: DURING accepts TODAY, YESTERDAY, LAST_7_DAYS, LAST_14_DAYS, LAST_30_DAYS, THIS_MONTH, LAST_MONTH, LAST_BUSINESS_WEEK, THIS_WEEK_MON_TODAY, LAST_WEEK_MON_SUN; for other periods use segments.date BETWEEN 'YYYY-MM-DD' AND 'YYYY-MM-DD'.

search_products

ChatGPT
Searches the products Clarmix holds for this account (Merchant Center data + Clarmix label + stored Google Ads metrics over the account default window + content scores + enrichment state). Use filters for labels, custom labels, brand, metric ranges or advanced conditions. sortBy can be slow on large result sets, so combine it with selective filters.

shopify_catalog_gaps

ChatGPT
Compares the live Shopify catalog with the products Clarmix has from Merchant Center: active Shopify variants not in Merchant Center (with a sample), top sellers of the last 30 days that are not in Merchant Center, variants in Merchant Center that Clarmix has not linked to Shopify, and Shopify variants missing a barcode (GTIN) or a product cost (COGS). Large stores are scanned for up to ~20 seconds, so counts can be partial (see scan.complete).

shopify_inventory

ChatGPT
Stock of the Shopify variants that are in Merchant Center, compared with sales velocity (30 / 90 days) and stored Google Ads spend: out-of-stock products still spending on ads, best sellers with low days of cover, overstock / dead stock (no sales in 90 days or too many days of cover), and products selling without ad spend, each with its Clarmix label. Totals include stock units and stock value at cost.

shopify_profitability

ChatGPT
For a period, joins Shopify sales and product costs (COGS) with live Google Ads Shopping / PMax spend per product: revenue (gross and net of discounts / returns / tax), units, gross profit (on net revenue when available), ad cost, ad conversion value, ROAS, POAS (gross profit ÷ ad cost) and profit after ads, per product, variant, vendor, product type or Clarmix label. Flags products that lose money after ad spend and reports ad spend that cannot be matched to a Shopify product. Profit figures need COGS ("Cost per item") set in Shopify and the same currency in Shopify and Google Ads.

shopify_query

ChatGPT
Runs a read-only Shopify Admin GraphQL query (API 2026-07) on the connected store for raw store details: a product with its variants, options, metafields or status, collections and their products, inventory items and their cost, locations, order statuses / items / discounts / refunds, shop settings. Results are raw Shopify data, not aggregated or joined with Google Ads. Query rules: query operations only; root fields: shop, products, product, productByIdentifier, productsCount, productVariants, productVariant, productVariantsCount, collections, collection, collectionByIdentifier, collectionsCount, inventoryItems, inventoryItem, locations, location, locationsCount, orders, order, ordersCount, metaobjects, metaobject, metaobjectDefinitions, sellingPlanGroups, publications; every list needs "first" (max 50), nested lists multiply (max ~2,500 items per query); use pageInfo { hasNextPage endCursor } and "after" to page. Customer and payment data (customer, email, phone, addresses, IPs, notes, payment details) is never available. Orders only go back 60 days before the store connected to Clarmix. Money is in the shop currency. Shopify GraphQL errors are returned so the query can be fixed.

shopify_sales

ChatGPT
Shopify orders for a period (stored by Clarmix from the connected store): orders, revenue, units, average order value, an optional day/week/month series, and the top products, variants, vendors, product types, collections or Clarmix labels by revenue or units, with net sales (after discounts, returns and cancellations, excluding tax and shipping). Shows how much revenue comes from products that are not in Merchant Center. Aggregates only, no customer data. Returns a "not connected" message with setup steps when the account has no active Shopify integration.

ads_account_performance

Claude

ads_change_history

Claude

ads_conversion_setup

Claude

ads_list_campaigns

Claude

ads_pmax_structure

Claude

ads_product_performance

Claude

ads_search_terms

Claude

ads_shopping_structure

Claude

check_label_delivery

Claude

get_account_overview

Claude

get_content_quality

Claude

get_enrichment_setup

Claude

get_labels_and_rules

Claude

get_product

Claude

gmc_account_issues

Claude

gmc_data_sources

Claude

gmc_product_issues

Claude

gmc_product_performance

Claude

list_accounts

Claude

run_gaql

Claude

search_products

Claude

shopify_catalog_gaps

Claude

shopify_inventory

Claude

shopify_profitability

Claude

shopify_query

Claude

shopify_sales

Claude

Similar Apps

Listed on MCP App Store

Listed on MCP App Store

Add this badge to your site or README. It links to this page.

HTML

<a href="https://mcp-app-store.keenethics-labs.com/app/clarmix"><img src="https://mcp-app-store.keenethics-labs.com/badges/listed-on-mcp-app-store.svg" alt="Listed on MCP App Store" width="220" height="44"></a>

Markdown

[![Listed on MCP App Store](https://mcp-app-store.keenethics-labs.com/badges/listed-on-mcp-app-store.svg)](https://mcp-app-store.keenethics-labs.com/app/clarmix)