Workflows endpoint

Workflows endpoint explains how this part of Flowline works and how to use it in a real workflow. This page is dummy content for testing the Cloudairy wiki: structure, navigation, search and rendering.

Note: Flowline is a fictional automation product. Section: Connect › Public API.

Prerequisites

• A Flowline account (cloud or self-hosted)
• Editor access to the project
• About 15 minutes

How it works

Every workflow starts with a trigger, passes items from node to node, and ends when the last node finishes. Workflows endpoint fits in the middle of that chain: it receives items, does its work, and hands the result to the next node. Use $json to read the current item and $node["Name"] to reach an earlier node.

Step-by-step

1. Open the workflow editor and select Add node.
2. Search for Workflows endpoint and add it to the canvas.
3. Fill in the required parameters (see the table below).
4. Select Test step to run it with sample data.
5. Check the output panel, then Save the workflow.

Parameters

Parameter	Type	Required	Description
Name	string	Yes	A label shown on the canvas
Mode	option	Yes	run once or run per item
Timeout	number	No	Seconds before the step fails (default 120)
Retry on fail	boolean	No	Retries up to 3 times with back-off

Example

# Workflows endpoint
docker run -it --rm \
  -p 5678:5678 \
  -e FLOWLINE_ENCRYPTION_KEY=change-me \
  -v flowline_data:/home/flowline/.flowline \
  flowline/flowline:latest

Warning: Changing workflows endpoint on a running production workflow takes effect on the next execution. Test in a copy first.

Advanced options

Set FLOWLINE_WORKFLOWS_ENDPOINT_LIMIT to cap how many items one execution processes. Leave it unset for no limit.

Scenario 1: Sync CRM contacts

This scenario chains 3 nodes. The trigger fires, the data is cleaned with Edit Fields, filtered with If, and sent onward. Typical runtime is 120 ms per item, and it handles up to 500 items per execution. Repeat the steps above, swapping the final node for the destination app.

• Input: 10 sample records
• Output: one message per matching record
• Failure handling: retry, then route to the error workflow

Scenario 2: Alert on failed payments

This scenario chains 4 nodes. The trigger fires, the data is cleaned with Edit Fields, filtered with If, and sent onward. Typical runtime is 240 ms per item, and it handles up to 1000 items per execution. Repeat the steps above, swapping the final node for the destination app.

• Input: 20 sample records
• Output: one message per matching record
• Failure handling: retry, then route to the error workflow

Scenario 3: Enrich new leads

This scenario chains 5 nodes. The trigger fires, the data is cleaned with Edit Fields, filtered with If, and sent onward. Typical runtime is 360 ms per item, and it handles up to 1500 items per execution. Repeat the steps above, swapping the final node for the destination app.

• Input: 30 sample records
• Output: one message per matching record
• Failure handling: retry, then route to the error workflow

Scenario 4: Archive old files

This scenario chains 6 nodes. The trigger fires, the data is cleaned with Edit Fields, filtered with If, and sent onward. Typical runtime is 480 ms per item, and it handles up to 2000 items per execution. Repeat the steps above, swapping the final node for the destination app.

• Input: 40 sample records
• Output: one message per matching record
• Failure handling: retry, then route to the error workflow

Scenario 5: Post daily reports

This scenario chains 7 nodes. The trigger fires, the data is cleaned with Edit Fields, filtered with If, and sent onward. Typical runtime is 600 ms per item, and it handles up to 2500 items per execution. Repeat the steps above, swapping the final node for the destination app.

• Input: 50 sample records
• Output: one message per matching record
• Failure handling: retry, then route to the error workflow

Scenario 6: Route support tickets

This scenario chains 8 nodes. The trigger fires, the data is cleaned with Edit Fields, filtered with If, and sent onward. Typical runtime is 720 ms per item, and it handles up to 3000 items per execution. Repeat the steps above, swapping the final node for the destination app.

• Input: 60 sample records
• Output: one message per matching record
• Failure handling: retry, then route to the error workflow

Scenario 7: Back up databases

This scenario chains 9 nodes. The trigger fires, the data is cleaned with Edit Fields, filtered with If, and sent onward. Typical runtime is 840 ms per item, and it handles up to 3500 items per execution. Repeat the steps above, swapping the final node for the destination app.

• Input: 70 sample records
• Output: one message per matching record
• Failure handling: retry, then route to the error workflow

Scenario 8: Translate messages

This scenario chains 10 nodes. The trigger fires, the data is cleaned with Edit Fields, filtered with If, and sent onward. Typical runtime is 960 ms per item, and it handles up to 4000 items per execution. Repeat the steps above, swapping the final node for the destination app.

• Input: 80 sample records
• Output: one message per matching record
• Failure handling: retry, then route to the error workflow

Troubleshooting

Symptom	Likely cause	Fix
Step shows No output	The previous node returned zero items	Pin sample data on the previous node
401 Unauthorized	Expired credential	Reconnect it under Credentials
Execution times out	Large payload	Split items with Split Out

Related

• Build your first workflow
• Expressions
• Error handling