What a flow chains together#

A flow is a visual canvas of blocks wired together port to port. You wire saved requests, logic and data blocks into one pipeline, then run the whole thing at once. Reach for a flow when a task takes more than a single request: chain one response into the next call, branch on a status code, or loop over a list.

When to use a flow#

A single request answers one question. A flow answers a sequence of them. Use one when you need to:

• Feed one request's response into the URL, headers or body of the next.
• Take a different path depending on what a response contains.
• Repeat a step over a list, a count or a generated set of values.

Every flow starts from one Start trigger. From there you add blocks and connect them. Data travels along the connections when you run the flow.

Open a new flow#

There are three ways to open a flow.

• Select the plus button next to the workspace tabs, then choose New Flow.
• Open a collection's menu and choose Add flow.
• Select a saved flow in the Collections panel to reopen it.

A new flow opens on its Start trigger, ready for you to add blocks.

Pages in this section#

• Workspace explains the block catalogue, canvas, inspector and execution log.
• Building a flow covers adding blocks, connecting ports and configuring them.
• Running a flow covers the run button and reading results.
• Execution log covers the timestamped event list.
• Inspector covers editing a block and reading its output.
• Saving a flow writes the flow into a collection.
• Exporting a flow copies the flow as JSON.
• HTTP Request block runs a saved request inside a flow.
• Flow block runs another flow as a sub-pipeline.
• Block reference is the floating panel that documents every block.

For the blocks themselves, see the reference pages for Actions, Logic, Looping, Data and Visualization.