Edit and run a pipeline

A Pipelab pipeline is a saved graph of plugin tasks and event triggers. The desktop editor is the main place to create and run one. In the command line, pipelab run runs a pipeline JSON file; it is different from pipelab workflow run, which runs a saved Release workflow.

Build the graph

Open a pipeline from a project, then add tasks from the task picker. The picker shows plugin tasks already available in the bundled build; it does not install or activate plugins. Add a trigger when the pipeline needs an event to start it; the available trigger nodes come from the loaded plugins. The built-in Manual event is one registered option.

Each task defines its own parameters, required fields, outputs, and possible platform constraints. Configure those fields in the task editor, save, then review any validation issues before running.

Variables and previous outputs

Pipeline variables have a name, description, and string value. The legacy pipeline evaluator resolves those values and task parameters when each action runs. Static JSON parameter values are read as JSON; other values are evaluated as JavaScript with variables, steps, and context available. A parameter evaluation error is logged and that parameter becomes an empty string, so check the task log if an advanced value is not behaving as expected.

This evaluator belongs to graph pipelines. Release workflows use separate run-variable and step-output references, described in Workflow inputs, outputs, and dependencies.

Execution and results

The graph executor walks action blocks in their saved array order. Disabled actions are skipped. Each successful action's output is stored under its block ID and may be used by later parameter evaluation. A task error stops the remaining graph; the editor shows task status, logs, and artifact paths. Use Cancel to abort a running graph.

Pipeline runs are recorded locally by default and artifacts are kept in the configured user-data area. The legacy Build History dialog requires an account with the Build History benefit. The separate Release workflow run list and details are covered in Runs and history.

Run a pipeline from the CLI

The CLI accepts a JSON pipeline file and can write the execution result to a file:

pipelab run ./pipeline.json --output ./result.json

The file must contain a graph or canvas.blocks list, and referenced plugin IDs must be available to the CLI. --variables accepts a JSON string that replaces the pipeline's saved variable list. There is no --dry-run option for pipelab run; use pipelab workflow run <id-or-name> --dry-run to inspect a Release workflow plan without executing its tasks.

See the CLI command reference for the full command syntax.