ferraus pipeline run

Run a pipeline file (YAML or JSON) with stage caching

Synopsis

ferraus pipeline run <input> [options] [-o <output>]

What it does

Run a pipeline file (YAML or JSON) with stage caching

Examples

Run every stage and write the outputs listed in the file

ferraus pipeline run job.yaml

Override one stage parameter for this run and get a JSON result

ferraus pipeline run job.yaml --set stages.tuned.args.strength=90% --json

Recompute every stage from scratch and replace existing outputs

ferraus pipeline run job.yaml --no-cache --overwrite

Options

FlagTypeDefaultMeaning
--settext—Override any field, e.g. stages.tuned.args.strength=90% Repeatable.
--no-cacheflag—Recompute every stage instead of reusing cached results

Global options

Every command also accepts the global options (-o, --json, --dry-run, --overwrite, --progress, ...). See Global options.

Exit codes

CodeMeaning
0Success
1Processing failed
2Usage or parameter error
3Input missing or unreadable
4Unsupported format or codec
8Cancelled
70Internal error (bug)

Common mistakes

  • Forgetting that outputs are never overwritten: pass --overwrite or a new -o path when re-running.

For models

  • MCP tool: ferraus_pipeline_run
  • Schema: ferraus schema pipeline run
  • Pipeline: { op: pipeline run, args: { ... } } (flag names with _ instead of -)
  • fx stage: no
{"command": "pipeline run", "defaults": {}}