AI workflow guides Your workspace
Find your way forward

Runs and debugging in Superagento

Learn to build AI agents, connect tools and
understand every workflow run.

Running/Runs and debugging1 min read

Runs and debugging

Every run keeps its receipts.

A run records each step's status, duration, summary, output and credit cost. A failed step stops its branch and is reported with the error the step raised, not a generic failure.

  • Test a single step rather than the whole graph while you are iterating.
  • Re-run the last payload after swapping a model to compare answers.
  • Read the per-account ledger on the billing page to see what each run was charged for.
A little help goes a long way.Your run history is a good place to find what happened at every step.
Take a closer look