# Execution logs (/docs/workflows/execution-logs)

> Find any workflow run, filter by result or type, and inspect what every node received and produced.

Open **Execution Logs** from the Workflow page, or **Executions** inside the
builder for one workflow only.

![The Execution Logs page](/docs-images/workflows/execution-logs.jpg)

1. Totals for completed, failed, waiting, and canceled runs. Waiting runs are paused on a Delay or Wait for Time node.
2. Search by run ID or workflow name.
3. **State** filter: Completed, Failed, Running, Waiting, Cancelled.
4. **Type** filter: Production (real events) or Test.
5. Date range.
6. and 7. Production and Test labels.
8. **Monitor** opens the run's live view. See [Read a run](/docs/workflows/test-and-publish#read-a-run).

Filter **State** to **Failed** once a day. Open each failed run, click the
red node, and read its Output to see what went wrong.

## What a call node records [#what-a-call-node-records]

When a run reaches a Voice AI Call, the node records the call's result.
Later nodes can use any of it.

![A run where the call was not completed](/docs-images/workflows/call-rejected-run.jpg)

This test called an invalid number, so the call failed and the run left
through **REJECTED** (5).

| Result | What it holds |
|---|---|
| Call status | Completed, failed, busy, or no answer. |
| Branch taken | Completed or Not Completed. |
| Call summary | A short AI summary of the conversation. |
| Conversation | The full transcript. |
| Extracted variables | Answers your agent captured, such as appointment time, interest level, or a callback request. |
| Duration and times | Call length, start time, and end time. |
| Recording | The call recording. |