Skip to main content
Data sources on a deployed build can be tested directly from Runtime: run a database operation or an SQL query and inspect the response, without wiring it into a workflow first. Use this to validate a data source after deploy or to investigate why an operation misbehaves in an environment.

​
Where to find it

1

Open the Builds list

In your project, go to Runtime and open the Builds list.
2

Open the build's test menu

On the build you want to test against, open the test actions menu.
3

Pick the test action for your data source

The available entry depends on the data source type configured in the build.
Data sourceTest action
FlowX Database, unmanaged MongoDBTest operation (this page)
Oracle, PostgreSQLTest query (this page)
RESTful systemTest endpoint
SOAP systemTest SOAP operation — see SOAP system
Knowledge BaseTest knowledge base — see Testing a knowledge base
The same menu also offers Test build (run a process from the build) and Test workflow (run an integration workflow) — see the Builds documentation for the full list.
Build test actions menu in Runtime, with Test operation and Test query entries
Testing data sources requires the data_source_read permission. Test runs are recorded in the audit log.

​
Testing SQL queries (Oracle and PostgreSQL)

Available starting with FlowX.AI 5.9.2
Test query opens a full-screen testing modal for the SQL data sources in the build:
  1. Pick a query from the selector in the modal header — queries are grouped by data source, and you can filter with Search query. Data sources without queries show No queries available.
  2. Review or edit the query in the Query Arguments editor (SQL, with autocomplete and a schema preview of the data source’s entities), and fill in any Input Parameters the query uses.
  3. Click Test. The Response pane shows the result as JSON, along with status, execution time, and size. Failures are shown with the database error message.
When opened from Runtime, the modal is read-only for the query definition — edits you make in the editor are for the test run only and cannot be saved to the build.
Queries run against the live database of that environment, and they are not restricted to read-only statements. An UPDATE, INSERT, or DELETE executes for real and may modify data — the modal shows the same warning. Test destructive queries only against non-production environments.

​
Testing database operations (FlowX Database and unmanaged MongoDB)

Test operation opens the equivalent modal for NoSQL data sources: pick an operation from the selector in the modal header (filter with Search operation), provide any input parameters, and click Test to see the operation’s response. The data source’s schema is available for reference from the modal.

​
Browsing documents

SaaS ·
Available on SaaS with FlowX.AI . This feature is live on managed (SaaS) deployments now. Self-hosted deployments receive it with the next LTS release.
For FlowX Database and unmanaged MongoDB data sources, the test modal gains a Documents tab alongside Test: browse the collection’s documents directly, filtering either visually with the Query Builder or by writing JSON filter and sort expressions, then click Run. Use it to check what an operation actually changed without writing a query for it.

FlowX Database

Collections, operations, and usage in workflows

Oracle database

Add an Oracle data source and define queries

PostgreSQL database

Add a PostgreSQL data source and define queries

Testing a knowledge base

The equivalent testing surface for knowledge bases
Last modified on July 15, 2026

Related topics

BuildsMCP Data SourcesSOAP SystemData mappersTesting a knowledge base