Skip to content
Talk to our solutions team

Atom reference

Every atom, with its parameters, defaults and worked examples. Each is reachable two ways — as a task in a flow and as a function in a script — and both are documented together.

GroupCovers
ExecutionShell, SSH, containers, Kubernetes, Python, scripts, HTTP
Files and transferFilesystem, archives, rsync, SCP, S3, Azure Blob
Data and templatingDatabases, JQ, Liquid, structured data, variables
Secrets and securityVault, secrets, hashing, encryption, certificates
Control and coordinationLocks, waits, suspend, throw, git, DNS
InfrastructureServices, processes, cron, ports, users, deployment

Three things save time on every page here.

Parameter names differ between the two surfaces, and not merely by case — a flow’s bucketname is a script’s bucket. Each entry gives both. See Tasks and functions.

Not every atom is available to every engine. A shell: task exists in an Automate flow and not in a data pipeline. Libraries has the mapping.

A flow is a graph, not a list. Multi-node examples here either declare routing with next: or run top-to-bottom with list: true. A node with neither is terminal — see Graphs and sequences.

If you know the kind of thing you need rather than its name, start from the library index:

  • Common — the atoms nearly every engine has
  • Infrastructure — hosts, services, processes
  • AI — models, embeddings, retrieval
  • Pipeline — readers, writers, transforms