Google Cloud Platform Toolkit
The Google Cloud Platform Toolkit is a workflow node. Its tools include Gcp Bigquery List Datasets, Gcp Bigquery List Tables, Gcp Bigquery Get Table Schema, and Gcp Bigquery Query, plus 9 more listed below. Create a connection before you use any of these tools. Type @ in a field when the value should come from an earlier step.
Authentication
Create a Google Cloud Platform connection before you use any of these tools.
- On the node, open Connection and choose Create Connection.
- Enter a Connection Label.
- Open Google Cloud Console, Service Accounts.
- Choose a service account, grant the roles this workflow needs, then open Keys, Add key, Create new key, and choose JSON.
- Paste that file into Service-account JSON.
- GCP Project ID, Default Region, and BigQuery Location are optional. Project ID defaults to the value in the JSON key.
- Save, then select the connection on the node.
Tools
Gcp Bigquery List Datasets
Runs the gcp bigquery list datasets tool.
| Field | Required | Description |
|---|---|---|
| Gcp Project ID | No | GCP project ID (e.g. 'my-gcp-project-123') to list datasets from. NOT the Odin workspace ID. OMIT this argument unless the user explicitly asked to query a different GCP project than the connection's default - the toolkit will fall back to the connection's gcp_project_id automatically. Default is null. Type @ to select a variable. |
| Max Results | No | Maximum number of datasets to return. Default is DEFAULT_PAGE_SIZE. |
Gcp Bigquery List Tables
Runs the gcp bigquery list tables tool.
| Field | Required | Description |
|---|---|---|
| Dataset ID | Yes | BigQuery dataset ID (e.g. 'analytics'). Type @ to select a variable. |
| Gcp Project ID | No | GCP project ID that owns the dataset (e.g. 'my-gcp-project'). NOT the Odin workspace ID. OMIT unless explicitly told to query a different GCP project than the connection's default. Default is null. Type @ to select a variable. |
| Max Results | No | Maximum number of tables to return. Default is DEFAULT_PAGE_SIZE. |
Gcp Bigquery Get Table Schema
Runs the gcp bigquery get table schema tool.
| Field | Required | Description |
|---|---|---|
| Dataset ID | Yes | Dataset that owns the table. Type @ to select a variable. |
| Table ID | Yes | Table ID (without project / dataset prefix). Type @ to select a variable. |
| Gcp Project ID | No | GCP project ID (e.g. 'my-gcp-project'). NOT the Odin workspace ID. OMIT unless explicitly told to query a different GCP project. Default is null. Type @ to select a variable. |
Gcp Bigquery Query
Runs the gcp bigquery query tool.
| Field | Required | Description |
|---|---|---|
| Query | Yes | Standard SQL query to run. Always use fully-qualified project.dataset.table references when crossing projects. Type @ to select a variable. |
| Max Results | No | Maximum rows to return from the query. Default is 100. |
| Use Legacy SQL | No | Set to true to use Legacy SQL. Standard SQL is recommended. Default is false. |
| Location | No | Dataset location (e.g. 'US', 'EU', 'asia-northeast1'). Defaults to the toolkit's configured location. Default is null. Type @ to select a variable. |
| Timeout Ms | No | Synchronous wait time in ms before the query is polled asynchronously. Default is 30000. |
| Gcp Project ID | No | GCP project ID that runs the query (billing project, e.g. 'my-gcp-project'). NOT the Odin workspace ID. OMIT unless explicitly told to bill a different GCP project. Default is null. Type @ to select a variable. |
| Allow Writes | No | Set to true to permit INSERT/UPDATE/DELETE/MERGE/DDL. Defaults to false so prompt-injection cannot trigger destructive statements. Default is false. |
| Max Bytes Billed | No | Cost cap (bytes). BigQuery rejects the job if its estimate exceeds this. Defaults to ~1 GiB; pass 0 to disable the cap (NOT recommended). Default is DEFAULT_BIGQUERY_MAX_BYTES_BILLED. |
Gcp Bigquery Insert Rows
Runs the gcp bigquery insert rows tool.
| Field | Required | Description |
|---|---|---|
| Dataset ID | Yes | Dataset containing the destination table. Type @ to select a variable. |
| Table ID | Yes | Destination table ID. Type @ to select a variable. |
| Rows | Yes | List of row dicts to insert. Keys must match the table schema. |
| Skip Invalid Rows | No | If true, valid rows will still be inserted when some rows fail validation. Default is false. |
| Ignore Unknown Values | No | If true, fields not present in the schema are silently dropped. Default is false. |
| Gcp Project ID | No | GCP project ID that owns the table (e.g. 'my-gcp-project'). NOT the Odin workspace ID. OMIT unless explicitly told to use a different GCP project. Default is null. Type @ to select a variable. |
Gcp Cloud Run List Services
Runs the gcp cloud run list services tool.
| Field | Required | Description |
|---|---|---|
| Region | No | Region/location to list Cloud Run services from (e.g. 'us-central1'). Defaults to the toolkit's region. Default is null. Type @ to select a variable. |
| Gcp Project ID | No | GCP project ID (e.g. 'my-gcp-project'). NOT the Odin workspace ID. OMIT unless explicitly told to query a different GCP project. Default is null. Type @ to select a variable. |
| Page Size | No | Maximum number of services to return. Default is DEFAULT_PAGE_SIZE. |
Gcp Cloud Run Get Service
Runs the gcp cloud run get service tool.
| Field | Required | Description |
|---|---|---|
| Service Name | Yes | Name of the Cloud Run service. Type @ to select a variable. |
| Region | No | Region of the service. Defaults to the toolkit's region. Default is null. Type @ to select a variable. |
| Gcp Project ID | No | GCP project ID (e.g. 'my-gcp-project'). NOT the Odin workspace ID. OMIT unless explicitly told to query a different GCP project. Default is null. Type @ to select a variable. |
Gcp Cloud Run Invoke Service
Runs the gcp cloud run invoke service tool.
| Field | Required | Description |
|---|---|---|
| URL | Yes | Fully-qualified Cloud Run service URL (e.g. 'https://my-svc-abc123-uc.a.run.app'). Type @ to select a variable. |
| Method | No | HTTP method (GET, POST, PUT, PATCH, DELETE). Default is POST. Type @ to select a variable. |
| Path | No | Path appended to the service URL. Defaults to '/'. Default is /. Type @ to select a variable. |
| Body | No | JSON request body (for POST/PUT/PATCH). Default is null. |
| Headers | No | Additional headers. Authorization is auto-injected. Default is null. |
| Query Params | No | Query string parameters. Default is null. |
| Use ID Token | No | Use a Google-issued ID token (required for private Cloud Run services). Set false to call public endpoints anonymously. Default is true. |
| Timeout | No | Request timeout in seconds. Default is DEFAULT_REQUEST_TIMEOUT. |
Gcp Cloud Run Invoke Function
Runs the gcp cloud run invoke function tool.
| Field | Required | Description |
|---|---|---|
| Function Name | Yes | Cloud Run Function (Gen 2) name. The toolkit looks up its trigger URL automatically. Type @ to select a variable. |
| Method | No | HTTP method (GET, POST, PUT, PATCH, DELETE). Default is POST. Type @ to select a variable. |
| Body | No | JSON body sent to the function. Default is null. |
| Query Params | No | Query string parameters appended to the trigger URL. Default is null. |
| Region | No | Region the function is deployed in. Defaults to the toolkit's region. Default is null. Type @ to select a variable. |
| Gcp Project ID | No | GCP project ID (e.g. 'my-gcp-project'). NOT the Odin workspace ID. OMIT unless explicitly told to invoke a function in a different GCP project. Default is null. Type @ to select a variable. |
| Timeout | No | Request timeout in seconds. Default is DEFAULT_REQUEST_TIMEOUT. |
Gcp Cloud SQL List Instances
Runs the gcp cloud sql list instances tool.
| Field | Required | Description |
|---|---|---|
| Gcp Project ID | No | GCP project ID (e.g. 'my-gcp-project'). NOT the Odin workspace ID. OMIT unless explicitly told to query a different GCP project. Default is null. Type @ to select a variable. |
| Max Results | No | Maximum number of Cloud SQL instances to return. Default is DEFAULT_PAGE_SIZE. |
Gcp Cloud SQL Get Instance
Runs the gcp cloud sql get instance tool.
| Field | Required | Description |
|---|---|---|
| Instance Name | Yes | Cloud SQL instance ID (e.g. 'analytics-prod'). Type @ to select a variable. |
| Gcp Project ID | No | GCP project ID (e.g. 'my-gcp-project'). NOT the Odin workspace ID. OMIT unless explicitly told to query a different GCP project. Default is null. Type @ to select a variable. |
| Include Databases | No | If true, also list databases on the instance. Default is true. |
Gcp Cloud SQL Execute Query
Runs the gcp cloud sql execute query tool.
| Field | Required | Description |
|---|---|---|
| Instance Connection Name | Yes | Full instance connection name (e.g. 'project:region:instance'). Type @ to select a variable. |
| Database | Yes | Database name to connect to. Type @ to select a variable. |
| Query | Yes | SQL query to execute. Read-only by default - pass allow_writes=True explicitly to permit DML/DDL. Type @ to select a variable. |
| Engine | No | Cloud SQL engine: 'postgres' or 'mysql'. Used to pick the driver. Default is postgres. Type @ to select a variable. |
| Max Rows | No | Maximum number of rows to return for SELECT queries. Default is 100. |
| Allow Writes | No | Set to true to permit INSERT/UPDATE/DELETE/MERGE/DDL. Defaults to false so prompt-injection cannot trigger destructive statements. Default is false. |
Gcp Custom API Call
Runs the gcp custom api call tool.
| Field | Required | Description |
|---|---|---|
| URL | Yes | Full Google API URL (e.g. 'https://compute.googleapis.com/compute/v1/projects/\{project\}/zones'). Type @ to select a variable. |
| Method | No | HTTP method (GET, POST, PATCH, PUT, DELETE). Default is GET. Type @ to select a variable. |
| Body | No | JSON request body for POST/PUT/PATCH. Default is null. |
| Query Params | No | Query string parameters. Default is null. |
| Headers | No | Additional headers. Authorization is auto-injected. Default is null. |
| Timeout | No | Request timeout in seconds. Default is DEFAULT_REQUEST_TIMEOUT. |
Related
Agents can use this integration from chat. See Google Cloud Platform.
See all integrations on the toolkits overview.