Does the pipeline run on a server?

No. The operations are the same local engines the panels use, so a pipeline runs in the browser on this machine.

What is in a template?

Operations and wiring only. The scientific values are set by you in the node inspector, so a template cannot smuggle in numbers you did not choose.

Why are some connections refused?

Because the handles are typed. An output that produces an alignment cannot be wired into an input that expects a primer pair, which is caught at the connection rather than at the run.

Record · experimental

Workflow Builder

The same operations the panels run, wired into a graph. Connections are typed, so an output can only be joined to an input that accepts it, and the values a step runs on are yours to set in the node inspector.

Open Workflow Builder Open the Workflow pipeline sample

Four real records and a restriction-map starter: wire tools into a DAG, fan it out over more than one molecule, Run, and save every step's output to timestamped project files

What it does

01Nodes drawn from the engine operation catalog, each a real capability rather than a placeholder
02Typed input and output handles, so a connection is only made where the types match
03Drag to connect, pan and zoom on the graph
04Node parameters set in an inspector, so the scientific values are yours
05Starter templates that supply the operations and wiring, not the values
06Run progress per node

What it runs on

Every node is a real operation from the `@ports/agent` catalog, and a connection is validated against the declared types before it is allowed. The panel runs on @ports/ligation, @ports/workflow.

Everything runs client-side. Files are parsed in the browser and projects are saved on this machine, so an unpublished sequence is never uploaded.

Questions

Does the pipeline run on a server?: No. The operations are the same local engines the panels use, so a pipeline runs in the browser on this machine.
What is in a template?: Operations and wiring only. The scientific values are set by you in the node inspector, so a template cannot smuggle in numbers you did not choose.
Why are some connections refused?: Because the handles are typed. An output that produces an alignment cannot be wired into an input that expects a primer pair, which is caught at the connection rather than at the run.

Also searched as

bioinformatics workflow builder · sequence analysis pipeline · node graph pipeline editor · no-code bioinformatics workflow · chain sequence tools · Galaxy alternative in the browser · sequence analysis pipeline without install · batch process sequence files · reproducible bioinformatics pipeline