Reference · Nodes
Core workflow nodes
Core nodes are the parts of a BusyBot workflow that belong to no outside service. They start a run (Manual Trigger, Webhook), decide where each item goes (If, Switch, Merge), reshape data between steps (Edit Fields, Filter, Code), repeat or pace work (Loop, Wait), and call any API that has no dedicated node (HTTP Request). Almost every workflow uses several of them, whichever integrations it connects, and most of them need no credential.
45 nodes — 43 actions and 2 triggers. Every page documents the shipped node: real parameters, real outputs, real error behaviour.
The building blocks of every workflow
The nodes in this category fall into five groups: Flow, Data Transformation, Helpers, Files and Triggers.
For flow, If sends each item to one of two outputs, Switch sends it down the first branch whose rule matches with a fallback for the rest, and Merge brings branches back together by appending, pairing by position or matching on a field. Loop repeats a chain of nodes for each item, a fixed number of times, or while a condition holds, and Wait pauses before passing items on.
For data, Edit Fields adds, renames and removes fields, Filter drops items that fail a condition, Item Lists sorts, limits and de-duplicates, and Aggregate collapses many items into one. When none of them can express the logic, Code runs JavaScript or Python.
For files, Extract from File turns CSV, spreadsheet, PDF, JSON, XML and ICS files into items, Convert to File goes the other way, and Compression zips and unzips. HTTP Request calls any URL with any method, which makes it the fallback when a service has no node of its own.
Triggers
Manual Trigger runs a workflow on demand. Webhook turns a workflow into an HTTP endpoint for any service that can send a callback. A saved workflow can also run on a schedule.
Example workflows
- Receive, route, export. Webhook receives data from an external service, Switch routes orders to different handlers by type, and Convert to File exports the results to a CSV file.
- Enrich within rate limits. Loop goes through each item and enriches it with an HTTP Request call, Wait pauses a few seconds between calls to respect the API’s rate limit, and Item Lists removes duplicate contacts at the end.
- Diff two exports. Extract from File parses two CSV files into row items, Compare Datasets finds the new, changed and deleted records, and HTML converts the result into a styled HTML table.
Actions
Triggers
Nodes that START a workflow when something happens.
Frequently asked questions
Which core nodes can start a workflow?
Manual Trigger starts a run when you click Execute and emits one item carrying a timestamp. Webhook gives the workflow a URL and starts it when an HTTP request arrives, with the body, headers, query parameters and method available to the nodes after it. A saved workflow can also run on a schedule, and the trigger nodes in the other categories start runs from events in those services.
Do core nodes need credentials?
Most do not. If, Switch, Merge, Filter, Edit Fields, Loop, Wait, Code and the file nodes run without one. The exceptions hold a secret or reach outside the workflow: HTTP Request can use basic, bearer, header or query authentication, Crypto takes an HMAC secret or a private key, and TOTP takes the shared secret it generates codes from.
How do core nodes work with AI and integration nodes?
They sit between them. A typical run fetches records with an integration node, keeps the ones that matter with Filter or If, sends each to an AI node such as OpenAI Structured Output, and brings the branches back together with Merge or Aggregate before writing the result somewhere. Nodes pass lists of items, so a node that receives a list usually runs once per item and looping is often implicit.
Last reviewed . Spotted something wrong? Tell us.