Guide · Agent skill
How to use the Application Integration Basics agent skill
Updated · written with AI from the project's own documentation and checked against our catalogue data
The Application Integration Basics skill gives an AI agent working knowledge of Google Cloud Application Integration, a managed service for building automated workflows that connect Google Cloud products and third-party business apps without managing servers. This guide shows you what the skill covers, how to install it from the ZIP on our page, how to ask for it in natural language, and what it needs and changes before you rely on it. It assumes you may never have used an agent skill before.
Application Integration Basicsgoogle/skills · Apache-2.0 Download the skill Quick start
- 1Download and unzip the skill. Download the skill ZIP from our page and unzip it into ~/.claude/skills/ for personal use, or into .claude/skills/ inside your project.
- 2Or upload it in the Claude apps. If you use the Claude apps, open Settings, go to Capabilities, then Skills, and upload the skill.
- 3Prepare your Google Cloud project. Confirm that the project has billing enabled and that gcloud is installed and signed in with working credentials.
- 4Ask a setup question in plain language. Describe the task, such as deciding between Application Integration and Cloud Run, or setting up a region. Include the project and target region.
- 5Review each proposed change. The agent will show the exact gcloud command, REST request or Terraform change. Read it and approve only what you intend to run.
- 6Check limits before you build. Ask the agent to check payload sizes, run-time limits and IAM setup so you know what the design can support.
What the skill makes the agent do
Agent skills are folders of instructions and reference files that an AI agent loads when a task matches them. This one is about setup and concepts for Application Integration. Once it is installed, the agent can help you decide whether this service fits a job or whether custom code is a better choice, walk through enabling APIs and setting up a region, explain IAM roles and run-as service accounts, and pick the right interface.
The skill also flags platform limits such as payload sizes and timeouts. It is guidance-focused: it asks for explicit approval before making any change to a cloud project.
- Compare Application Integration with custom code on Cloud Run or Cloud Functions
- Explain regions, triggers, tasks, data mappings and Integration Connectors connections
- Walk through enabling APIs and provisioning a region
- Explain IAM roles and run-as service accounts
- Choose between the REST API, the Python discovery client, Terraform and the ADK toolset
- Point out limits such as payload sizes and timeouts
How to install the skill
Download the ZIP for this skill from our page and unzip it. The skill folder must end up in one of two places. For a personal setup, use ~/.claude/skills/. For a single project, use the .claude/skills/ folder inside that project.
If you use the Claude apps instead of a local setup, you can upload the skill under Settings, then Capabilities, then Skills.
- Personal install: unzip into ~/.claude/skills/
- Project install: unzip into .claude/skills/ inside your project folder
- Claude apps: Settings → Capabilities → Skills, then upload the skill
How to trigger the skill with natural requests
You do not need to type a special command. Describe the task in plain language that matches what the skill covers, such as setting up Application Integration, enabling its APIs, provisioning a region, configuring IAM, or checking quotas and timeouts. The agent loads the skill when the request fits.
Requests about building, publishing or invoking integrations, or about diagnosing failed runs, are handled by other skills, so ask for those directly if that is what you need.
- Should I use Application Integration or Cloud Run for this workflow?
- Walk me through enabling the APIs for Application Integration in my project.
- Which IAM roles does a run-as service account need for my integration?
- Set up Application Integration in europe-west1 and explain the steps first.
- What payload size and timeout limits apply to API-triggered integrations?
Choose the right interface
The skill explains several ways to work with Application Integration. There is no gcloud integrations command group and no Google-managed remote MCP server for controlling the service. The REST API is the primary control and execution interface.
The Python discovery client covers the same v1 API through googleapiclient. Terraform manages infrastructure such as region setup and auth configs, but not integration workflow versions. The ADK ApplicationIntegrationToolset exposes published API-trigger integrations or connector actions as tools for ADK agents. The Cloud console provides the visual editor and approval handling.
- REST API with curl: full v1 surface, authenticated with a bearer token from gcloud
- Python discovery client: install google-api-python-client and google-auth, then authenticate with Application Default Credentials
- Terraform (hashicorp/google): region setup, auth configs and connectors infrastructure only
- ADK ApplicationIntegrationToolset: exposes published integrations or connector actions to ADK agents
- Cloud console: visual editor, region setup and OAuth consent flows
gcloud auth print-access-tokenWhat you need before you start
You need a Google Cloud project with billing enabled, the gcloud CLI for enabling APIs and reading IAM policies, and Google Cloud credentials through gcloud auth or Application Default Credentials. Python with google-api-python-client and google-auth, Terraform with the hashicorp/google provider, or Google ADK are optional, depending on the interface you choose.
The skill assumes cloud administration knowledge and changes to billing-linked resources. It is not a fit for users who do not have a Google Cloud project.
- Google Cloud project with billing enabled
- gcloud CLI installed and signed in
- Credentials via gcloud auth or Application Default Credentials
- Optional: Python libraries, Terraform with hashicorp/google, or Google ADK
gcloud services listSafety notes: scripts and network
This skill does not run scripts, but it does use the network. It may run read-only checks freely, such as listing enabled services, reading the project IAM policy, and sending GET requests to the Application Integration and Integration Connectors APIs.
Before any state change, the agent must show you the exact command or payload and wait for your approval. That covers enabling APIs, provisioning a region, modifying IAM policies, creating or publishing integrations, and changing global gcloud or credential state. Enabling CMEK for a region cannot be undone, so review that step carefully. The skill also says it never bundles scripts that delete integrations, versions, auth profiles or connections unattended.
- Read-only checks run without asking
- Enabling APIs, provisioning regions and changing IAM always need your approval
- Installing Python packages or changing gcloud config or credentials needs your approval
- CMEK for a region cannot be undone
- Destructive scripts are not bundled unattended
CLOUDSDK_METRICS_REQUEST_ATTRIBUTION="agent-skills/1.0 (skill:application-integration-basics)" gcloud <command> [flags]Tips for getting better answers
Name your target region and project when you ask. Application Integration is regional, and provisioning one region does not provision another. If you have no preference, the skill suggests us-central1 as a well-supported default, after checking that the region is listed as supported.
Ask for the limits early. Payload sizes, the 2-minute synchronous and 10-minute asynchronous run limits, and the fact that Call REST Endpoint reaches only the public internet all affect whether this service fits. Ask the agent to compare the options before you start building.
- Always name the project and region in your request
- Check the region against the supported locations list first
- Ask for limits before designing the workflow
- Review each proposed command before approving it
What people use it for
- Deciding whether a CRM-to-ERP sync should use Application Integration or a custom Cloud Run service
- Enabling the required APIs and provisioning a region in a new Google Cloud project
- Setting up a run-as service account with the right IAM roles for an integration
- Choosing between Terraform and the REST API for managing region setup in a team repository
- Checking payload and timeout limits before designing an event-driven workflow
Frequently asked questions
How do I install the Application Integration Basics agent skill?
Download the skill ZIP from our page and unzip it into ~/.claude/skills/ for personal use, or into .claude/skills/ inside a project. In the Claude apps, you can instead upload it under Settings, then Capabilities, then Skills.
Is the Application Integration Basics skill free?
The skill is published under the Apache-2.0 licence. Google Cloud Application Integration itself is a paid Google Cloud service that requires a project with billing enabled, so check Google Cloud pricing separately for your usage.
What do I need to use this skill?
You need a Google Cloud project with billing enabled, the gcloud CLI, and credentials through gcloud auth or Application Default Credentials. Python libraries, Terraform with the hashicorp/google provider, or Google ADK are optional depending on the interface you pick.
Does the skill run scripts or access the network?
It does not run scripts, but it does use the network. Read-only checks run freely, while enabling APIs, provisioning regions, changing IAM and similar actions require your explicit approval first.
Can this skill build or troubleshoot my integrations?
No. It covers setup and concepts only. Building, publishing and invoking integrations are handled by a separate authoring skill, and diagnosing failed or stuck runs is handled by a separate troubleshooting skill.
How does Application Integration compare with Cloud Run or Cloud Functions?
Application Integration suits SaaS connections, event-driven workflows and visual JSON transformation without containers. Cloud Run or Cloud Functions suit sub-second latency, custom binary protocols, payloads over 30 MB, and runs longer than the 2-minute synchronous or 10-minute asynchronous limits.
New agent skills, repos and AI tools — every week
The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.
Our verdict
Use this skill if you are setting up Application Integration in a Google Cloud project for the first time, or deciding whether a workflow belongs on Application Integration or on custom code. Pick something else if you need to build, publish or debug integrations, or if you do not already have a Google Cloud project with billing enabled.
Download the skill