What Test run does
Test run evaluates the flow against the most recent matching example, checks path conditions, and previews configured actions. It does not send Slack messages, emails, webhooks, or X posts.
Transform, Decode, and NFT metadata execute to provide outputs. Agent only previews its request; it does not generate a response. Later steps that depend on Agent or a published X post cannot be fully rendered in a dry run.
A preview validates the data and configuration it can see, not future delivery, service availability, or every possible event. Test runs do not consume Tailflow credits.
Find a real run
Open Runs in the flow designer to inspect that flow. Open Activity for the account-wide run feed and select a run for details.
Completed means execution finished successfully. Failed means a step or condition failed. Skipped means a path or step did not execute; a nonmatching path is expected to be skipped. Running means execution is still in progress.
Read the evidence
Inspect the path condition, action status, what was sent, and the returned result. Start with the first failure; later skipped actions often follow from it.
A completed request is evidence of acceptance at that step, not proof that a human read a message. Temporary failures can leave a flow Active; configuration faults or repeated failures can stop it with Error status. Check both the run and the flow status.
History is not permanent
Detailed sent and returned content is retained for a shorter period than the run summary. Older runs may keep status and usage information after their detailed content or original event is unavailable. Capture the run link and relevant details promptly when asking for help.