Transform MCP exposes 21 tools grouped into five families. Tool names use underscores, for example environments_list, to satisfy MCP client naming rules.
List tools return a { data, next, limit } envelope where pagination applies. Pass startingFrom from a previous response's next field to fetch the next page.
For setup, see Connect to Transform MCP. For ID formats, see Find Coalesce IDs.
Manage deployment Environments in your organization.
environments_list
List deployment Environments in your organization.
| Input | Type | Required | Description |
|---|
detail | boolean | No | Include full detail, including Storage Locations and health, rather than summaries |
limit | integer | No | Page size, 1–500 (default 100) |
startingFrom | string | No | Pagination cursor from a previous response's next |
project | string | No | Filter to a single Project ID |
environments_get
| Input | Type | Required | Description |
|---|
environmentID | string | Yes | The Environment ID |
environments_create
| Input | Type | Required | Description |
|---|
name | string | Yes | Environment name |
project | string | Yes | Project ID the Environment belongs to |
oauthEnabled | boolean | Yes | Whether OAuth is enabled for the Environment connection |
description | string | No | Environment description |
connectionAccount | string | No | Account name for the Environment connection |
accessUrl | string | No | Optional access URL for the Environment connection |
defaultStorageMapping | string | No | Default Storage Mapping name |
mappings | object | No | Storage Location mappings keyed by location name |
environments_update
| Input | Type | Required | Description |
|---|
environmentID | string | Yes | The Environment ID to update |
name | string | No | Environment name |
description | string | No | Environment description |
connectionAccount | string | No | Account name for the Environment connection |
accessUrl | string | No | Access URL for the Environment connection |
defaultStorageMapping | string | No | Default Storage Mapping name |
oauthEnabled | boolean | No | Whether OAuth is enabled |
mappings | object | No | Storage Location mappings |
environments_delete
| Input | Type | Required | Description |
|---|
environmentID | string | Yes | The Environment ID to delete |
Manage Projects, which group Environments and Workspaces.
projects_list
projects_list is not paginated upstream. The response always fits on one page. The next field is null.
| Input | Type | Required | Description |
|---|
includeWorkspaces | boolean | No | Include nested Workspace data for all Projects |
includeJobs | boolean | No | Include nested Job data for all Workspaces in the Projects |
projects_get
| Input | Type | Required | Description |
|---|
projectID | string | Yes | The Project ID |
includeWorkspaces | boolean | No | Include nested Workspace data |
includeJobs | boolean | No | Include nested Job data for all Workspaces |
projects_create
| Input | Type | Required | Description |
|---|
platformKind | string | Yes | snowflake, databricks, bigquery, or fabric |
name | string | Yes | Project name |
description | string | No | Project description |
gitRepoURL | string | No | Git repository URL |
autoPreviewEnabled | boolean | No | Whether sample data is fetched automatically after running a Node |
projects_update
| Input | Type | Required | Description |
|---|
projectID | string | Yes | The Project ID to update |
platformKind | string | No | snowflake, databricks, bigquery, or fabric |
name | string | No | Project name |
description | string | No | Project description |
gitRepoURL | string | No | Git repository URL |
autoPreviewEnabled | boolean | No | Whether sample data is fetched automatically after running a Node |
projects_delete
| Input | Type | Required | Description |
|---|
projectID | string | Yes | The Project ID to delete |
Inspect run history and results.
| Tool | Purpose | REST equivalent |
|---|
runs_list | List runs with filters and pagination | Get Runs |
runs_get | Get one run by numeric run ID | Get Run |
runs_results | Get per-run results for a deploy or refresh | Get Run Results |
runs_list
| Input | Type | Required | Description |
|---|
limit | integer | No | Maximum runs to return, 1–1000 |
startingFrom | number or string | No | Pagination cursor from a previous response's next |
orderBy | string | No | Field to sort by. Set explicitly when using startingFrom |
orderByDirection | string | No | Sort order |
projectID | string | No | Filter to a single Project ID |
runType | string | No | Filter by run type |
runStatus | string | No | Filter by run status |
environmentID | string | No | Filter to a single Environment ID |
detail | boolean | No | Include full run detail rather than summaries |
runs_get and runs_results
| Input | Type | Required | Description |
|---|
runID | integer | Yes | Numeric run ID. The scheduler API calls this value runCounter. |
runID must be the numeric run counter, for example 42, not the UUID in a run URL. See Find Coalesce IDs.
Inspect Nodes in an Environment or Workspace.
| Tool | Purpose | REST equivalent |
|---|
nodes_list_environment | List Nodes deployed in an Environment | Get Nodes |
nodes_get_environment | Get one Environment Node | Get Node |
nodes_list_workspace | List Nodes in a Workspace | Get Workspace Nodes |
nodes_get_workspace | Get one Workspace Node | Get Workspace Node |
nodes_list_environment
| Input | Type | Required | Description |
|---|
environmentID | string | Yes | The Environment ID |
detail | boolean | No | Include full Node detail rather than summaries |
limit | integer | No | Page size, 1–500 (default 100) |
startingFrom | string | No | Pagination cursor from a previous response's next |
skipParsing | boolean | No | Skip parsing column references and updating sources |
nodes_get_environment
| Input | Type | Required | Description |
|---|
environmentID | string | Yes | The Environment ID |
nodeID | string | Yes | The Node ID |
skipParsing | boolean | No | Skip parsing column references and updating sources |
nodes_list_workspace
| Input | Type | Required | Description |
|---|
workspaceID | string | Yes | The Workspace ID |
detail | boolean | No | Include full Node detail rather than summaries |
limit | integer | No | Page size, 1–500 (default 100) |
startingFrom | string | No | Pagination cursor from a previous response's next |
skipParsing | boolean | No | Skip parsing column references and updating sources |
nodes_get_workspace
| Input | Type | Required | Description |
|---|
workspaceID | string | Yes | The Workspace ID |
nodeID | string | Yes | The Node ID |
skipParsing | boolean | No | Skip parsing column references and updating sources |
Trigger and manage refresh runs. These tools call the scheduler API and can start warehouse compute.
runs_start
Trigger a refresh or DML run on a deployed Environment using the Workspace stored connection.
| Input | Type | Required | Description |
|---|
environmentID | string | Yes | Target Environment ID |
jobID | string | No | Job to scope the refresh to |
includeNodesSelector | string | No | Selector for Nodes to include |
excludeNodesSelector | string | No | Selector for Nodes to exclude |
parallelism | integer | No | Number of Nodes to process in parallel |
forceIgnoreWorkspaceStatus | boolean | No | Run even if Workspace status would normally block it |
confirmRunAllNodes | boolean | No | Must be true when neither jobID nor includeNodesSelector is set |
runs_retry
| Input | Type | Required | Description |
|---|
runID | integer | Yes | Numeric run ID to re-run |
forceIgnoreWorkspaceStatus | boolean | No | Re-run even if Workspace status would normally block it |
runs_cancel
| Input | Type | Required | Description |
|---|
runID | integer | Yes | Numeric run ID to cancel |
environmentID | string | Yes | Environment ID the run belongs to |
runType | string | No | Type of run to cancel |
runs_status
| Input | Type | Required | Description |
|---|
runID | integer | Yes | Numeric run ID to look up |
Terminal statuses are completed, failed, and canceled. Non-terminal statuses are waitingToRun and running.
What's Next?