Every run, on the record
Reliability is mostly visibility. GoRunner records every execution in enough detail to debug it, replay it, and prove what happened.
- ✓New email receivedgmail0.02s
- ✓Classify and draftgorunner-ai0.31s
- ✓Route by intentbranchtook path A: billing0.00s
- ✗Create casesalesforceREQUIRED_FIELD_MISSING: Priority (fixed in draft v15)0.69s
- ○Alert #supportslack–
A real log to poke at. Open the failed run, then replay it.
Step-level logs
Each run records every step's input, output, duration, and retry history. When something fails at 3am, you read what happened instead of reconstructing it.
Replay it, two ways
Take any historical run and retry it with the exact trigger data it saw: against the latest version, or against the version it originally ran on. Lineage is recorded either way.
Retries that know better
Exponential backoff with per-step retry and timeout overrides, dead-letter handling for runs that never recover, and durable suspend for approvals and delays: parked runs release the worker.
Alerts where you live
When a run fails, your error workflow fires with the run link in its payload. Route it to Slack, PagerDuty, email, or any of the 600+ pieces. Alerting that is itself a flow.
Usage you can predict
Run counts and durations per flow, surfaced daily. You see the bill coming, which is more than most automation tools can say.
Audit-grade history
Who deployed what and when, which version each run used, and what every AI step decided. Exportable, and retained on your terms.
Trust, but verify everything.
The free plan keeps seven days of run history. Paid plans keep as much as you need.
Free plan · 1,000 runs a month · self-host or cloud