Google Ads API agent skills are portable, self-contained modules of Google Ads API-specific knowledge, instructions, and workflows. Google Ads API agent skills work with AI assistants that support the open Agent Skills standard, including Google Antigravity and Claude Code.
These agent skills are designed to help AI assistants understand Google Ads API best practices, configure tools, and execute workflows with higher accuracy and lower token cost. For a list of available skills, see Available skills.
Get started
Install the agent skills appropriate for your development environment.
Install Google Ads API agent skills
In most cases, Google Ads API agent skills can be installed in your preferred editor with a single command:
Antigravity
Google Ads API agent skills are included as one of the Build with Google integration bundles for Antigravity. You can enable this bundle for global-level access at two points:
- During onboarding: Select the checkbox for the Google Ads API stack.
- In Settings: Navigate to Settings > Customizations. Under Build with Google Plugins, click Customize, then click Download for the Google Ads API integration.
If you prefer project-level access, run the following command in your project's directory:
npx skills add google/skills --agent=antigravity
Claude Code
claude plugin marketplace add google/skills
claude plugin install google-ads@skills
Codex
npx skills add google/skills
Cursor
You can install Google Ads API agent skills directly by adding the skill's GitHub
repository endpoint (for example, https://github.com/google/skills) in
your editor's settings under custom rules. For more details, see the
Cursor rules documentation.
Other agents
For other AI assistants supporting the open Agent Skills standard, you can install the skills using NPX:
npx skills add google/skills
Update Google Ads API agent skills
You can get all available updates and clean up out-of-date skills by running the
update command:
Antigravity
Integrations enabled through Build with Google are managed and updated automatically by the editor.
If you installed agent skills using the npx command, you can get all
available updates by running the update command:
npx skills update --all
Claude Code
claude plugin update google-ads@skills
Codex
npx skills update --all
Cursor
npx skills update --all
Other agents
For other AI assistants supporting the open Agent Skills standard, you can update all active skills using NPX:
npx skills update --all
Use Google Ads API agent skills
AI assistants are designed to use skills automatically whenever they detect that a skill's description matches your current request. Globally installed skills remain dormant in unrelated projects unless your prompt context matches Google Ads API tasks.
Skills can also be manually invoked by typing / in the agent chat and
searching for the skill name.
Core components of a skill
Each agent skill is a specialized package that provides the AI assistant with the following components:
- Specialized instructions: Detailed prompt instructions injected into the context window to guide the assistant with generating GAQL queries, configuring OAuth credentials, and handling common error codes.
- Best practices: Standard performance, security, and integration recommendations defined by the Google Ads API DevRel team.
- Automation scripts: Executable scripts (such as for environment validation or workspace setup) included in the skill packages to execute tasks locally.
Benefits of using skills
Agent skills extend an assistant's capabilities with specialized knowledge on demand without the overhead of loading bulky instructions into every conversation.
Reduced token costs
Conventional AI integrations often load massive amounts of documentation upfront, which consumes significant tokens and increases session costs. Agent skills use progressive disclosure to minimize this overhead:
- The AI assistant initially only checks the skill's name and description to see if it is relevant to your request.
- The skill's full instructions and references are only added to the active context window when the assistant determines they're necessary for your specific task, saving context space.
AI-guided implementation
Instead of manually searching through documentation, you can describe your intent in natural language. Skills guide your AI assistant to perform tasks such as the following:
- Generate a GAQL query to retrieve campaign performance metrics from my account.
- Write a Python script using client libraries to create a new campaign.
- Retrieve campaign conversions and diagnostic data using the Google Ads API.
For a full list of the available skills, see Available skills.
Use skills alongside the Google Ads API MCP server
Agent skills are designed to complement the Google Ads API Model Context Protocol (MCP) server:
- Google Ads API MCP server: Designed for AI-assisted development workflows, enabling AI assistants to interact with your Google Ads API resources, campaigns, and metrics programmatically.
- Google Ads API agent skills: Provide prompt instructions and recommended practices that guide an assistant to perform tasks through token-efficient progressive disclosure. They instruct the agent on how to use tools like the MCP server and client libraries effectively.
Available skills
The Google Ads API team provides a suite of skills for building with the Google Ads API:
| Skill | Description |
|---|---|
google-ads-api-mcp-setup |
Assists with setting up and configuring the Google Ads API Model Context Protocol (MCP) server, simplifying authentication and configuration in local environments. |
google-ads-api-quickstart |
Guides you through setting up your developer environment, configuring credentials, and making your first API call to the Google Ads API. |