Where to look
Every trigger fire creates a run you can inspect. Runs show up alongside your other executions and history, and each run links back to the trigger that started it, the company or contact involved (if any), and the messages and tool calls Swan made along the way. To see a single trigger’s runs, use the “View logs” shortcut on the trigger page’s Activity card — it opens the Desk filtered to that trigger (old/triggers/{triggerId}/logs links redirect there too). The Desk lists completed and in-progress runs only; to find a trigger’s failed or cancelled runs, ask Swan in chat (for example, “show me the failed runs of my inbound form trigger this week”).
You can search past runs by keyword or by meaning — the search understands what you’re describing even if you don’t remember the exact wording used in the run, and it also covers past conversations, not just trigger runs. You can filter by company, status, and date range.
Who sees what
- Admins see the runs of every trigger in the organization.
- Everyone else sees the runs of org-wide triggers (the ones without a single owner) plus the runs of their own personal triggers. You won’t see another member’s personal trigger runs unless you’re an admin.
Run status
Each run has a status: pending, in progress, completed, failed, cancelled, or awaiting approval (when a step needs your sign-off before Swan continues). A run that finishes but only partially completes its instructions is still marked completed — check the run’s summary for what it actually got done versus what it couldn’t.Website Changes checks
A Website Changes poll records a check for every enrolled company the data provider can match, even when no downstream agent runs. Its summary shows whether the website-change data was missing, unchanged, excluded, filtered out, or matched. You can filter history to the Website Changes trigger and search by company domain or summary. Only a Matched check continues into the trigger’s instructions. An enrolled domain that the provider cannot match to a business has no per-company check record for that poll.Common failure reasons
- Insufficient credits — your organization had fewer than the 10 credits required to start when the event fired. The run doesn’t start; the trigger stays enabled and picks back up once enough credits are available. See how triggers work.
- Trigger disabled — the trigger was turned off between the event being created and the run starting.
- Disconnected integration — a tool the run needed (a CRM, inbox, or connected app) lost its connection. Reconnect it from connected apps or your inboxes and the next event will run normally.
- Timed out — a company-scoped job or sub-agent run that wasn’t waiting (for approval, a delay, or other sub-agents) made no progress for an hour and was cancelled with
[swan] Timed Out. Scheduled, form, and org-level webhook runs are not cancelled this way. You can ask Swan in chat to do the same work, or let the trigger fire again from its normal source.