Google has open sourced the Mannequin Context Protocol (MCP) server, which exposes read-only entry to the Google Adverts API for brokers and LLM functions. The repository googleads/google-ads-mcp implements an MCP server in Python that gives two instruments: search (GAQL queries for advert accounts) and list_accessible_customers (enumeration of buyer assets). This contains setup with pipx, Google Adverts developer token, OAuth2 scopes (https://www.googleapis.com/auth/adwords), and Gemini CLI / Code Help integration with normal MCP consumer configuration. This undertaking is labeled “experimental.”
So why is it necessary?
MCP has emerged as a standard interface for connecting fashions to exterior methods. By transport a reference server for the Adverts API, Google reduces integration prices for LLM brokers that require marketing campaign telemetry, finances pacing, and efficiency diagnostics with out the necessity for customized SDK glue.
How does it work? (Developer perspective)
Protocol: MCP standardizes the “instruments” that fashions can invoke with typed parameters and responses. The Adverts MCP server advertises instruments which might be mapped to Google Adverts API operations. MCP shoppers (reminiscent of Gemini CLI/Code Help) detect and invoke them throughout a session. Authentication and scoping: Allow the Google Adverts API in your cloud undertaking, receive a developer token, and configure default credentials or the Adverts Python consumer in your utility. The required scope is Adwords. For administrator account hierarchy, set the login buyer ID. Shopper connection: Add a ~/.gemini/settings.json entry pointing to the MCP server name (pipx run git+https://github.com/googleads/google-ads-mcp.git google-ads-mcp) and go the credentials through atmosphere variables. Then run the question through Gemini’s /mcp or by prompting for issues like campaigns and efficiency.
ecosystem indicators
Google’s servers come amid rising adoption of MCP throughout distributors and open supply shoppers, strengthening MCP as a sensible path to agent-to-SaaS interoperability. For PPC and rising groups experimenting with agent workflows, reference servers are a low-friction approach to validate LLM-assisted QA, anomaly triage, and weekly studies with out granting write permissions.
Necessary factors
Google has open sourced the read-only Google Adverts API MCP server and launched two instruments: search (GAQL) and list_accessible_customers. Implementation particulars: Python undertaking (googleads/google-ads-mcp) on GitHub, Apache-2.0 license, marked Experimental. Set up/run through pipx and configure OAuth2 with https://www.googleapis.com/auth/adwords scope (growth token + elective login buyer ID). It really works with MCP-compatible shoppers (reminiscent of Gemini CLI / Code Help), so brokers can subject GAQL queries and analyze advert accounts via pure language prompts.
conclusion
In truth, Google’s open supply Google Adverts API MCP Server supplies groups with a standards-based, read-only path via which LLM brokers can run GAQL queries in opposition to advert accounts with out the necessity for customized SDK wiring. The Apache licensed repository is marked as experimental, exposes search and list_accessible_customers, and integrates with MCP shoppers reminiscent of Gemini CLI/Code Help. For manufacturing use, you must take into account OAuth scopes (adwords), developer token administration, and information leakage warnings within the README.
Take a look at our GitHub web page and technical weblog. Be at liberty to go to our GitHub web page for tutorials, code, and notebooks. Additionally, be happy to observe us on Twitter. Additionally, do not forget to hitch the 100,000+ ML SubReddit and subscribe to our e-newsletter. cling on! Are you on telegram? Now you can additionally take part by telegram.
Michal Sutter is a knowledge science professional with a grasp’s diploma in information science from the College of Padova. With a powerful basis in statistical evaluation, machine studying, and information engineering, Michal excels at remodeling complicated datasets into actionable insights.
🙌 Observe MARKTECHPOST: Add us as your most well-liked supply on Google.
