Record Import Step

pre_upsert_records takes a list of records, usually the body of an http_get, and creates or updates one Promptless CPT Pages record for each. It needs Promptless CPT Pages 0.9.0 or later (0.10.0 or later for featured_image_url); without it the step fails with dependency_missing. The full workflow around it is in Importing records.

Settings

Setting	Required	What it does	Default
post_type	Yes	The record type’s slug, registered in Promptless CPT Pages.	—
source	Yes	A short, permanent key for the system the records come from, such as recdesk.	—
records	Yes	The list, usually "{{ steps.fetch.body.items }}". An object of records is treated as a list.	—
map	Yes	How each record becomes a CPT Pages record; see below.	—
expect_min_records	No	Fail if fewer records arrive.	1
max_failure_ratio	No	Fail if more than this share of records cannot be written, from 0 to 1.	0.1
missing_upstream	No	keep or draft: what to do with records of this source the list no longer contains.	keep

The map

The map is filled in once per record, with the current record available as {{ item.<field> }}. Other paths (vars, steps, now()) work too.

Key	Required	What it sets
external_id	Yes	The record’s id in the other system. With post_type and source, it is the record’s identity.
title	Yes	The title.
content, excerpt	No	The body and the excerpt.
status	No	The WordPress status, such as publish or draft. A new record without one is published; an existing one keeps its status.
fields	No	An object of CPT Pages field keys and values.
taxonomies	No	An object of taxonomy slugs and terms: a list, or text separated by commas.
featured_image_id	No	A Media Library attachment id.
featured_image_url	No	An image address. It is downloaded once per address and reused on later runs.

"map": {
  "external_id": "{{ item.id }}",
  "title": "{{ item.name }}",
  "fields": { "event_start": "{{ item.startDate }}", "location": "{{ item.facility.name }}" },
  "taxonomies": { "program_category": "{{ item.category }}" }
}

What happens to each record

Result	When
Created	No record of this type has this source and external_id.
Updated	One exists and the mapped values changed. Only the keys in the map are written.
Unchanged	One exists and nothing in the mapped values changed. Nothing is written.
Failed	external_id or title came out empty, or CPT Pages refused the record. Listed in failed; the other records carry on.

After every record, the step checks its guards:

• fewer than expect_min_records records arrived: the step fails with upstream_shape before writing anything;
• more than max_failure_ratio of them failed: the step fails with upstream_shape; the records already written stay.

Only when both guards pass and missing_upstream is draft are published records of this source that were not in the list switched to draft. Records are never deleted.

Output

Field	What it holds
received_count	Records in the list.
created_count, updated_count, unchanged_count, failed_count, drafted_count	Counts by result.
created_ids, updated_ids, unchanged_ids, drafted_ids	The post ids.
failed	Each failure’s index, external_id and error.
warnings	Map paths missing on some records (with counts), and taxonomy or image problems.

The run history records the map as written, not filled in.

Available since FlowMint Workflows 0.8.0.

Related pages

• Importing Records — Build a scheduled workflow that fetches another system's records and creates or updates Promptless CPT Pages records without duplicates, and keeps a broken feed from emptying the site.
• HTTP Steps — Call another system's API with http_get, http_post or http_request — settings, how bodies are encoded, what the output holds, and how status codes become errors.
• Variables and Expressions — The {{ }} placeholder syntax, every value a workflow can read (data, labels, entry, steps, vars and more), the built-in functions, and how conditions are written.