Configuration basics
An automation configuration describes what to process, where to exchange it, and how to map its fields. The API reference describes the request used to create or update an automation; this guide explains the configuration concepts.
Connection
The connection object describes how to reach the other system. Depending on the integration, this can be a file transfer server, an HTTP API, or email. The required settings depend on the connection type. Keep passwords and tokens in protected parameters rather than copying them into examples or sharing them in messages.
File configuration
file_configs is an array. Each entry describes a file or payload and its processing rules. Common settings include:
| Setting | Purpose |
|---|---|
name or regex |
Selects a file by name or pattern. Requirements depend on the connection and workflow. |
delimiter |
Sets the character between columns; the default is a comma. |
headers |
Describes import columns or whether an export includes a header row. |
field_map |
Connects SureDone fields to the other system's fields. |
field_run |
Makes additional mapped fields available for processing without importing them as output fields. |
identifier |
Chooses the SureDone field used to match an incoming product or order. The defaults are guid for products and oid for orders. |
create, update, remove |
Choose which import operations the automation performs. |
search |
Selects records using SureDone's search syntax. |
vendor_actions, suredone_actions |
Apply conditions and functions to records. See Actions. |
In field_map, SureDone field names are the keys and the other system's field names are the values. For example:
{
"field_map": {
"guid": "supplier_sku",
"stock": "quantity_available"
}
}
For an import, this maps supplier_sku to the SureDone identifier and quantity_available to stock. Confirm the supplier's column names and your matching rules before running it.
Matching and filtering
For priority matching, identifiers can contain an ordered list of field names; it overrides identifier when supplied. The automation checks nonempty identifier values in order for a match. If none matches, it uses the first nonempty value when creating a new record. Identifier-targeting actions can run before matching, with the limitations described in Actions.
Use a file-level search or file-level filters, rather than both: a search generates and replaces the file-level filters. Conditions within actions are separate. In SureDone search syntax, spaces join terms as AND, while parentheses group alternatives; literal AND and OR are not operators. See the product search reference for syntax and examples.
Checking changes
For imports, diff_update defaults to true unless the run overrides it. With diff_fields, an update is considered when a listed field changes. Product exports use a separate per-item snapshot mechanism when diff_fields is set: products with no prior snapshot are included, and a successful export advances the snapshot. A full export or diff_update: false includes all matched products and advances the baseline; debug runs neither read nor advance these snapshots. Avoid listing response writeback fields in diff_fields, because the writeback can make an item appear changed again. A failed run can require follow-up: if processing stops after advancing a snapshot but before its response writeback completes, a later diff run may not regenerate the missing writeback until a listed field changes.
Settings and supported combinations vary by automation type and connection. Validate the configuration and test a representative sample before enabling a schedule.