CSV Connector

Overview

Utility connector for working with CSV content inside a runbook. It parses an in-memory CSV string into structured data, including an optional header row and positional data rows. The input is normalized to UTF-8 before parsing, making it suitable for processing CSV content from previous actions, HTTP responses, or file attachments.

Actions

Parse CSV

Parses an in-memory CSV string into an optional header row and positional rows. Blank lines are skipped, and ragged rows are preserved.

Use case: Convert a CSV payload from an upstream action into structured rows that a runbook can iterate over.

Input Parameters

Parameter	Type	Required	Default	Description
content	String	Yes	-	The CSV text to parse.
headers	Boolean	No	true	Treat the first row as a header row.
col_sep	String	No	,	Field delimiter.
quote_char	String	No	"	Quote character.
declared_encoding	String	No	UTF-8	Fallback encoding used only when the input has no byte-order mark.

Example Input

{
  "content": "name,city\nAlice,NYC\n"
}

Output

Field	Type	Description
header	String[]	The first row when headers is true; otherwise empty.
rows	Array[]	Parsed data rows. Each row is returned as a list of string values. Ragged rows are preserved.
truncated	Boolean	Indicates whether the input contained more rows than the connector returned.
source_encoding	String	The detected or declared encoding before conversion to UTF-8.
lossy_encoding	Boolean	Indicates whether any characters were replaced during UTF-8 conversion.

Example Output

{
  "header": [
    "name",
    "city"
  ],
  "rows": [
    [
      "Alice",
      "NYC"
    ]
  ],
  "truncated": false,
  "source_encoding": "UTF-8",
  "lossy_encoding": false
}

Error Handling

The action fails with a clear error message when the CSV input cannot be parsed, an invalid option is supplied (such as an unknown declared_encoding, an empty col_sep, or a multi-character quote_char), or when a row exceeds 16,384 columns, which typically indicates an incorrect delimiter.

Ragged rows are preserved with their original length. If character conversion to UTF-8 results in data loss, the lossy_encoding output is set to true.

Best Practices

• Combine the header output with each row to enable name-based column access.
• The connector returns a maximum of 65,536 rows. Check the truncated output to determine whether additional rows were omitted.
• Verify the lossy_encoding output before processing text originating from non-UTF-8 sources.
• Use the parsed rows with Loop or For Each actions to process CSV records efficiently within a runbook.

Common Use Cases

• Parse CSV files retrieved from HTTP endpoints.
• Process CSV attachments from email connectors.
• Import data into business applications from CSV exports.
• Transform CSV reports into structured data for downstream actions.
• Iterate over CSV records to automate bulk operations.