How to use the Google Ads API Quickstart agent skill

What does the Google Ads API Quickstart skill make the agent do?

When installed, the agent can guide you from zero to a working Google Ads API request. It covers getting your developer token and OAuth2 credentials, choosing a client library or raw REST calls, setting up the environment, and generating a script that retrieves your campaigns.

It also helps diagnose common setup errors such as USER_PERMISSION_DENIED, login_customer_id issues and DEVELOPER_TOKEN_NOT_APPROVED.

• Supported tracks: Python, Java, .NET, PHP, Ruby or Perl client libraries, or direct REST HTTP requests.
• Credential walkthrough: developer token, OAuth2 client ID and secret, and OAuth2 refresh token.
• Script generation: a 'retrieve campaigns' example for your chosen track.

How do you install the Google Ads API Quickstart skill?

Download the ZIP of the skill from the directory page and unzip it. Put the resulting folder in one of these locations, depending on how you use your agent.

For personal use across projects, unzip it into ~/.claude/skills/. For a single project, unzip it into that project's .claude/skills/ folder.

If you use the Claude apps instead, upload the skill under Settings → Capabilities → Skills.

• Personal install: ~/.claude/skills/
• Project install: .claude/skills/ inside your project
• Claude apps: Settings → Capabilities → Skills, then upload

What do you need before you start?

The skill assumes you already have a Google Ads Manager account with API Center access and a developer token. You also need a Google Cloud project with the Google Ads API enabled and OAuth2 client credentials. The skill walks you through creating these, but it does not create accounts for you.

You need a language runtime such as Python, Java, .NET, PHP, Ruby or Perl. The skill looks up current minimum versions online and falls back to versions stored inside the skill if the lookup fails.

Outbound HTTPS access to Google documentation and GitHub is needed for the live lookups.

• Google Ads Manager account with API Center access and a developer token
• Google Cloud project with the Google Ads API enabled and OAuth2 client credentials
• A language runtime (Python, Java, .NET, PHP, Ruby or Perl)
• Outbound HTTPS access to Google documentation and GitHub

How do you trigger the skill with a natural request?

You do not need a special command. Describe what you are trying to do in plain language, and the agent will load the skill when your request matches it. The skill's description says it applies when you ask how to get started with the Google Ads API, need credentials or a developer token, want a quickstart script, or hit errors such as USER_PERMISSION_DENIED or DEVELOPER_TOKEN_NOT_APPROVED.

• How do I get started with the Google Ads API in Python?
• I need to set up a Google Ads developer token. Where do I find it?
• Write a quickstart script that retrieves my campaigns using the Java client library.
• I'm getting USER_PERMISSION_DENIED when I call the Google Ads API. What should I check?
• Should I use the .NET client library or raw REST calls for my Google Ads integration?

How do you set up the client library once you have credentials?

The agent gives you install steps for your chosen language. For example, the .NET reference installs the official package with NuGet, and the Perl reference installs the client with cpanm. Your credentials go into a configuration file or into code, depending on the library.

For Perl, the library looks for a googleads.properties file. It checks an explicit path passed to Client->new first, then the GOOGLE_ADS_CONFIGURATION_FILE_PATH environment variable, and then a default location.

What does a typical session look like?

Ask the agent for a quickstart in your language. It first resolves current API and runtime versions, then asks you to gather the five authentication parameters, including the developer token and OAuth2 credentials. The skill creates the OAuth consent screen and client steps for the Google Cloud Console, then guides you through generating the refresh token with the gcloud CLI.

If you hit an error, paste the full message into the chat. The skill includes troubleshooting guidance for USER_PERMISSION_DENIED, login_customer_id problems and DEVELOPER_TOKEN_NOT_APPROVED.

• Developer token from the API Center, signed in with a Manager account
• OAuth2 client created as a Desktop App, saved locally as client_secrets.json
• Test user added to the OAuth consent screen, using the email you log into Google Ads with

What safety notes should you know about?

The skill runs scripts and uses the network, so review anything it generates before running it. Its own instructions say it resolves versions from live sources and falls back to stored ones when that fails.

Do not use a pending developer token against live accounts. The skill warns that a token that is not yet approved can only target Google Ads test accounts, and a production account call will fail with DEVELOPER_TOKEN_NOT_APPROVED.

Keep client_secrets.json and your credentials private. Do not paste them into shared chats or commit them to public repositories.

• Runs scripts and uses the network: check generated code before running it.
• Pending tokens work only with test accounts.
• Treat your OAuth client file and tokens as secrets.

What tips help you get the most from it?

Name your language and whether you want a client library or REST in your first message, so the agent picks the right reference. Tell it if you need a specific API version; otherwise it resolves the latest stable one.

If you are choosing between a client library and direct REST calls, ask for a comparison before you commit to one. Keep the generated script in a version-controlled project so you can see what changed when an error appears.

• Say your language and track (library or REST) in the first message.
• Paste the exact error text when troubleshooting.
• Ask for a comparison of client library and REST before choosing.