Released 2026-06-04. You can now test every agent type from the editor — not just interactive chat agents.
Added
- A Test tab on the agent editor, for every agent type. Interactive agents keep the developer chat panel. Webhook, schedule, and playground agents get a Run test button: supply a sample input (the JSON body a webhook would post, or a sample trigger payload; scheduled agents take none) and trigger a run from the editor. The panel shows the run's outcome and links straight to the full run so you can inspect every step.
Good to know
- A test run is a real run: it uses your agent's live configuration and can send real emails, write real data, trigger real workflows, and pause for real approvals. It uses your credits and counts toward your daily run limit, and appears under Runs (where it can be retried) — the test panel says this up front.