--- title: Agent activity logs slug: agent-platform/docs/agent-activity-logs docTags: createdAt: 2026-08-12T15:37:07.000Z --- The **Agent Activity Logs** tab unifies execution activity for [specialized agents](docId\:WyA1gsk6tWr84Hum7a1Nx) and [workflow agents](docId:2HD0fUsIsfdek516VtN9l) in Agent Platform. Monitor recent executions, troubleshoot failures, and cancel runaway executions in one place. ## Access the Agent Activity Logs tab To open the Agent Activity Logs tab from the Agents page, go to **Agents > Agent Activity Logs**. ![screenshot of the Agents page where the Agent Activity Logs tab is selected](https://api.archbee.com/api/optimize/BrbvVORKJ5iIeSMZye5kI/9ED5WPkNbLbOfQ1k-DPdK_access-the-agent-activity-logs-tab.png) To open the Agent Activity Logs filtered to a specific workflow agent, click **More** > **View Logs** for that agent. See [Workflow agent logs](docId\:cLVkgspflAkgfjWgc9w9V) for details. ## Filter executions The **Agent Activity Logs** tab includes filters that narrow the list of executions in the table. ![screenshot of the Agent Activity Logs tab where the Agent Type, Status, Triggered By, and Date Range filters are visible](https://api.archbee.com/api/optimize/BrbvVORKJ5iIeSMZye5kI/Zavgpsq2pA4CmSC9_weY8_filter-executions.png) - **Agent Type filter** – Filter executions by agent type. - **Status filter** – Filter executions by status. See [Execution statuses](docId\:rHBhJmLiQvmx9789rso0n) for descriptions of each status. - **Triggered By filter** – Filter executions by the user who triggered them. - **Date Range filter** – Filter executions by timeframe. ### Search by agent name Enter an agent name in the **Search by agent name** field. Results update as you type. ![screenshot of the Agent Activity Logs tab where an agent name is entered in the Search by agent name field](https://api.archbee.com/api/optimize/BrbvVORKJ5iIeSMZye5kI/mNUSHmc7SZCug_UclDtIw_search-by-agent-name.png) ## Configure columns Show, hide, and reorder columns in the table to control which information displays. 1. Click **Configure Columns**. ::Image[]{src="https://api.archbee.com/api/optimize/BrbvVORKJ5iIeSMZye5kI/wZ86XI1wNtWmmkL-83k6l_configure-columns.png" alt="screenshot of the Configure Columns panel where column toggles are visible" size="100" isUploading="false" width="800" height="428" darkWidth="800" darkHeight="428" showCaption="false" indent="1"} 2. Toggle a column on or off. 3. Drag a column to reorder it. ## Table columns The table includes the following columns: - **Type** – An icon indicating the agent type (specialized or workflow). - **Execution Time** – The date and time the execution started. - **Agent** – The agent name and ID. - **Trigger** – The trigger type that started the workflow execution. See [Workflow agent triggers](docId\:pYeMYR_1QCBBfc6rGEwlR) for descriptions of each trigger type. - **Triggered By** – The user who triggered the execution. - **Status** – The current execution status. See [Execution statuses](docId\:rHBhJmLiQvmx9789rso0n) for descriptions. - **Credits** – The approximate credit usage for the execution. - **Evaluation Score** – The quality score for the execution. This column displays a value only for executions of agents that have preferred output examples. - **Duration** – The total time the execution took to run. ## Private executions When an agent has **Private executions** enabled, its executions still display in the table, but only the user who triggered each run can open the contents. - A lock icon marks every private execution. - Private executions that you did not trigger are greyed out, and you cannot open them. The tooltip reads **This is a private execution — only its owner can open it.** - Your own private executions display normally, with the lock icon. See [Create a specialized agent](docId\:tZPJWDyE1GSB-epLsUh5S) to enable private executions for an agent. ## Cancel multiple executions Cancel multiple executions at once instead of stopping each one individually. This is useful when a workflow triggers more executions than expected or when many executions queue up. 1. Select the checkbox for each execution to cancel. To select all executions, select the checkbox in the column header. 2. Click **Stop executions**. ::Image[]{src="https://api.archbee.com/api/optimize/BrbvVORKJ5iIeSMZye5kI/Ry-wNCAiXg3YkOlueCyy9_cancel-multiple-executions.png" alt="screenshot of the Agent Activity Logs table where executions are selected and the Stop executions option is visible" size="100" isUploading="false" width="800" height="377" darkWidth="800" darkHeight="377" showCaption="false" indent="1"} 3. Click **Cancel Executions** to confirm. Mark cancels only executions with **Running** or **Pending** statuses. When you cancel, any step in progress finishes and no further steps start. ## Execution statuses The **Status** column displays the current state of each execution. Each status badge also has a reason that explains what happened and why. Hover over a status badge to see its reason. Examples include waiting for execution capacity and a user stopping the execution. Use the reason to find the cause of an unexpected result and decide what to do next, without guessing or contacting support. ![screenshot of the Status column where the pointer rests on a status badge and its reason displays](https://api.archbee.com/api/optimize/BrbvVORKJ5iIeSMZye5kI/LL_P1PEpB02BELUDBfYg-_agent-activity-logs-execution-status.png) The following table lists each status, its reasons, and what each reason means. Descriptions that end with "(workflows only)" apply only to workflow agents. | Status | Reason | What it means | | -------------- | --------------------------------- | ----------------------------------------------------------------------------------------------------------------------------------- | | **Pending** | **Queued** | The execution is waiting in line to start. | | **Pending** | **Waiting for a free slot** | All capacity is busy. The execution starts when a slot opens. | | **Pending** | **Deferred for priority** | The execution gave up its slot so higher-priority work could go first. | | **Pending** | **Rate limited** | Too many executions started recently. The execution waits until the limit resets. | | **Running** | **Started** | The execution is working. | | **Running** | **Resumed** | The execution continued after a retry, a pause, or an answer to its question. | | **Paused** | **Waiting on input** | The execution asked a question or is waiting on an external service. | | **Completed** | **Success** | The execution finished normally. | | **Failed** | **Timeout** | The execution took too long and was cut off. | | **Failed** | **Error** | The execution hit an error it could not recover from. | | **Failed** | **Validation failed** | The inputs did not pass checks, so nothing ran. | | **Failed** | **Output rejected** | The result was reviewed and judged unusable (workflows only). | | **Failed** | **Invalid setup** | The workflow configuration is invalid, so the workflow never started (workflows only). | | **Stopped** | **Unusual behavior detected** | The system halted the execution before it finished because its behavior looked anomalous. | | **Stopped** | **No tools available** | The system blocked every tool the execution tried to use, so the execution could not continue. | | **Stopped** | **Blocked by content moderation** | Content moderation flagged the content, so the system never sent it to the model. | | **Stopped** | **Blocked by input rules** | The agent's own rules rejected the request. | | **Stopped** | **Security threat detected** | A scan flagged a file for suspicious content. | | **Stopped** | **No matching path** | A condition checked every option, none matched, and there was no fallback (workflows only). | | **Stopped** | **Too many iterations** | The execution hit the maximum allowed iterations (workflows only). | | **Stopped** | **Too many steps** | The execution hit the maximum allowed steps (workflows only). | | **Stopped** | **Gave up** | The workflow planner, the component that decides which step runs next, determined that it could not make progress (workflows only). | | **Cancelled** | **Stopped manually** | A person stopped the execution. | | **Cancelled** | **Disconnected** | The browser tab closed or the connection dropped during the execution. | | **Superseded** | **Replaced by newer run** | A newer message in the same conversation took over. | | **Superseded** | **Replaced by retry** | The system retried the failed execution and abandoned this one. | | **Superseded** | **Replaced by resume** | The system resumed a cancelled execution and abandoned this one. | | **Unknown** | **Not recorded** | The system did not capture the outcome. | ## Related articles - [Specialized agent logs](docId\:nt3nlnNvyD_-F82soWgev) – View detailed logs for specialized agent executions. - [Workflow agent logs](docId\:cLVkgspflAkgfjWgc9w9V) – View detailed logs for workflow agent executions. :::hint{type="info"} **Note** If you use Opti ID, administrators can turn off generative AI in the Opti ID Admin Center. See [Turn generative AI off across Optimizely applications](docId\:X5U0kKD2azV5cIdNqwGNj). :::