Step Types Overview

Learn about all available step types in SyncMyOrders.

How to read this page

Each step is a building block in a workflow. Use these descriptions to choose the right step for the job—no coding required.
“Map data” means either pick a value that already exists (for example, data.order.id or steps.check_stock.outputs.available) or type a fixed value (text/number/yes-no).
Most flows start with Start and end with Finish; the rest route, repeat, or call into agents and subflows.

What it does: Marks where the scenario begins. It receives the incoming payload defined by your scenario’s input schema.
When to use: Always present as the first node; no configuration needed.
Typical use: Immediately follow it with a step that reads from data.* (the incoming values) or sets defaults with an Agent/Transform step.

What it does: Calls a capability exposed by an agent (examples: HTTP request, Shopify “create product,” CSV “parse file,” OpenAI “extract fields”).
How to use: Select the agent, pick a capability/action, choose a connection if required, then map each input field to a value or reference.
When to use: Anytime you need to talk to an external system or apply a ready-made transform.
Typical use cases: Make an API call, clean/transform data, format text, enrich with AI, upload/download files.

What it does: Checks a condition and routes the flow based on true/false.
How to use: Build the condition with operators like equals, contains, greater/less than, is empty, etc. Reference values (for example, data.total or steps.lookup.outputs.status) or set literal values to compare against.
When to use: To skip, retry, or reroute when data meets certain criteria.
Typical use cases: Stop if an order is missing a field, send VIP orders down a special path, or only call an API if stock is available.

What it does: Chooses one of many paths based on a value, similar to a multi-option router.
How to use: Pick the value to check, then add cases with match types (equals, contains, greater/less than, ranges, starts/ends with). Optionally add a default path when nothing matches.
When to use: When you have more than two outcomes or need clear routing by category/state.
Typical use cases: Route by order status, marketplace/channel, file type, country/region, or error code.

What it does: Repeats a subflow for every item in an array.
How to use: Point it to the list (for example, data.items), then build the subgraph of steps to run for each item. Choose sequential (one by one) or parallel, and whether to stop on first failure.
When to use: To process collections like order lines, rows in a CSV, files in a batch, or API results pages.
Typical use cases: For each line item, reserve inventory; for each product, normalize fields then upsert; for each record, call an external API.

What it does: Invokes a child scenario, optionally pinning to a specific version or always using the latest.
How to use: Pick the child scenario and version, then map the parent data into the child’s inputs.
When to use: To reuse shared logic (like “normalize customer data” or “post tracking updates”) across multiple flows.
Typical use cases: Fan out to a dedicated subflow per channel, call a shared enrichment routine, or trigger a downstream integration you maintain separately.

What it does: Marks the end of the scenario and maps values you want to return.
How to use: Map outputs from previous steps or the incoming data into the final result object.
When to use: Always present as the last node in any path to produce a clean, predictable response.
Typical use cases: Return a success flag, IDs created upstream, transformed objects, or a summary of per-item results from a Split.