> ## Documentation Index
> Fetch the complete documentation index at: https://docs.mage.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Create block

`POST /api/pipelines/:pipeline_uuid/blocks`

Create a new block in one or more pipelines.

<ParamField query="name" type="string" required>
  Human readable block name. This is used to generate the UUID, e.g. `restless dust`. Names will be slugified to a safe-uuid.
</ParamField>

<ParamField query="type" type="string" required>
  Type of block:
  `callback`, `chart`, `conditional`, `custom`, `data_exporter`, `data_loader`, `dbt`, `extension`, `global_data_product`, `markdown`, `scratchpad`, `sensor`, `transformer`
</ParamField>

<ParamField query="color" type="string">
  Block color displayed by the frontend for `custom` blocks. All other block types have predefined colors.
</ParamField>

<ParamField query="config" type="object">
  An optional parameter specifying a block configuration file For example, `{data_source: "s3"}` will create an s3 data loader.
</ParamField>

<ResponseField name="configuration" type="object">
  Miscellaneous configuration settings for the block.

  <Expandable title="properties">
    <ResponseField name="data_provider" type="string">
      Database or data warehouse for the SQL block to connect to.
    </ResponseField>

    <ResponseField name="data_provider_database" type="string">
      Database name to use when saving the output of the SQL block.
    </ResponseField>

    <ResponseField name="data_provider_profile" type="string">
      Profile target for the dbt block.
    </ResponseField>

    <ResponseField name="data_provider_schema" type="string">
      Schema name to use when saving the output of the SQL block.
    </ResponseField>

    <ResponseField name="data_provider_table" type="string">
      Table name to use when saving the output of the SQL block.
    </ResponseField>

    <ResponseField name="export_write_policy" type="string">
      Whether to `replace` the existing table of the SQL block output, `append`, or
      raise an error and `fail`.
    </ResponseField>

    <ResponseField name="use_raw_sql" type="string">
      Toggle writing raw SQL in the block. Read more [here](/guides/blocks/sql-blocks#using-raw-sql).
    </ResponseField>
  </Expandable>
</ResponseField>

<ParamField query="content" type="string">
  The content of block formatted as a string. Use `/n` for newlines. Omitting will create an empty block.
</ParamField>

<ParamField query="extension_uuid" type="string">
  UUID of extensions, currently [Great Expectations](/development/testing/great-expectations) is the only extension supported.
</ParamField>

<ParamField query="language" type="string">
  The block language:
  `python`, `sql`, `r`, `yaml`, `markdown`
</ParamField>

<ParamField query="pipelines" type="[string]">
  Array of strings containing the pipeline uuids for which the block is a member.
</ParamField>

<ParamField query="priority" type="integer">
  The 0-indexed position of the block in the pipeline, e.g. 0 is the first block location.
</ParamField>

<ParamField query="upstream_blocks" type="[string]">
  An array of strings representing the uuids of upstream blocks (the blocks on which the new block depends). For example, `["a", "b"]` will create a block that depends on blocks `a` and `b`.
</ParamField>

<ParamField query="uuid" type="string">
  The unique id of the block. This is generated from `name`, so it's not necessary to populate this field.
</ParamField>

<RequestExample>
  ```bash Request theme={"system"}
  curl 'http://localhost:6789/api/pipelines/broken_wave/blocks?api_key=zkWlN0PkIKSN0C11CfUHUj84OT5XOJ6tDZ6bDRO2' \
  -H 'Accept: application/json, text/plain, */*' \
  -H 'Authorization: Bearer 018Cc8ydDUa6YWoLad2oMdw0EKdWWJuyg85D02GN1ww' \
  -H 'Content-Type: application/json' \
  --data-raw '
    {
    "block": {
        "name": "silent fire",
        "priority": 2,
        "configuration": {
            "data_source": null
        },
        "language": "python",
        "type": "data_exporter"
    },
    "api_key": "zkWlN0PkIKSN0C11CfUHUj84OT5XOJ6tDZ6bDRO2"
  }
  ' \
  --compressed
  ```
</RequestExample>

<ResponseExample>
  ```json Response theme={"system"}
  {
      "block": {
          "content": null,
          "name": "silent fire",
          "priority": 2,
          "config": {
              "data_source": null
          },
          "language": "python",
          "type": "data_loader",
          "configuration": {},
          "upstream_blocks": []
      },
  }
  ```
</ResponseExample>