Flowmage v4.11.4 – Patch Notes

New: Recruition Node

Flowmage can now talk to Recruition — reading, creating and updating candidates, vacancies and the rest of the API v2 surface.

• Recruition is hosted per customer, so one Tenant field is all the connection needs: acme becomes https://acme.recruition.nl/api/v2, and a full hostname or URL is accepted just as well
• The dialog shows the resulting base URL live while you type, so a typo is visible before you run anything
• Authenticate with email + secret — a session token is fetched automatically on every run — or paste an access token yourself; tokens expire after an hour, so email + secret is the default and the right choice for scheduled flows
• Thirteen resources, grouped as Recruitment (candidates, vacancies, procedures), Records (memos, tasks, files, users) and Reference data (branches, educations, job functions, companies, states, salary ranges)
• The operation dropdown only offers what the endpoint actually supports — List, Get, Create, Update, Delete — so no request can be built that Recruition will refuse
• Standalone endpoints are available as their own operations: the authenticated user (/me), identity (/identity), and candidate owners — list them, replace the whole list, or remove a single owner
• Candidates and vacancies get a search-as-you-type picker for the ID field; the smaller reference tables get a dropdown of actual records, so you rarely have to know an ID by heart
• Workflow variables such as @previousNode.data.id@ work in every field, including the ID and the credentials
• Test node runs the configured request against Recruition straight from the dialog and shows the response
• Node output is split into data, meta, raw, request and status, so a later node can read a clean list without digging through the API envelope

Recruition Node: Filters, Pagination and Request Bodies

Each resource carries its own field list, and the dialog uses it everywhere.

• The filter builder offers exactly the fields that resource accepts as a query parameter, instead of a free-text guess
• Candidates and vacancies additionally accept custom fields, filtered as custom_field_data->field_name
• Sorting per field, ascending or descending, on the same field list
• Pagination has two modes: Single page with page and per page, or All pages, which follows the API until it runs out of pages or hits the max pages you set
• An all-pages run reports how many pages it fetched, how many records it collected, and whether it stopped early against the maximum
• Request bodies can be built as key/value rows or written as raw JSON; numbers, true, false, null and JSON values are sent as that type, everything else as text
• Rows with an empty value are left out of the request entirely — Recruition rejects an empty string the same way it rejects a missing field, so a half-filled row no longer breaks the whole call. Use the JSON mode to send an empty value on purpose
• Update states plainly that it changes the single record with the ID above, never the whole collection
• A Raw request operation covers anything not in the list: any method, any path, your own body
• Advanced holds extra query parameters, custom headers, timeout, continue on error and in-depth logging; the node explains which headers it sets for you and what your own headers override
• Errors come back readable — a failed login, a missing tenant, a timeout or an API error all name what went wrong, and with continue on error the flow proceeds with the error in the node output

Other Improvements

Minor bug fixes and performance optimizations