Agent Automation
Overview
The Automations page displays all executions of published Agents within the AI Command Center.
Each entry on the Automations page represents an Agent Automation, which is a single execution of a published Agent.
The Automations page allows users to monitor, control, and audit running or completed automation executions.
From this page, you can:
- Monitor execution progress and activity status
- View logs, notifications, and runtime data
- Suspend, resume, cancel, or redirect execution flow
- Diagnose execution errors and licensing issues
How to Access
From the AI Command Center, navigate to: Reports → Automations
Automation Views
The Automations page provides status-based filters that allow users to quickly view automations according to their execution state.
| View | Description |
|---|---|
| All | Displays all automation executions regardless of their current status. |
| Pending | Shows automations that are waiting to start execution. |
| Initiated | Displays automations that have been successfully created and are preparing for execution. |
| Running | Shows automations that are currently executing. |
| Suspended | Displays automations that have been manually or automatically paused. |
| Cancelled | Shows automations that were stopped before completion. |
| Completed | Displays automations that finished successfully. |
| Faulted | Shows automations that failed due to errors during execution. |
Filtering Automations by Creation Date
You can filter automations by their creation date using the calendar filter located in the Created At column.
| Option | Description |
|---|---|
| Today | Displays automations created today |
| Yesterday | Displays automations created yesterday |
| Last 24 Hours | Displays automations created in the last 24 hours |
| Last 7 Days | Displays automations created in the last 7 days |
| Last 30 Days | Displays automations created in the last 30 days |
| This Month | Displays automations created in the current month |
| Last Month | Displays automations created in the previous month |
| Custom Range | Allows filtering using a custom date range |
Available Actions
The following actions are available for each automation.
| Action | Description |
|---|---|
| View | Opens the automation execution details page. |
| Audit | Displays the audit history for the automation. |
Initiating an Agent Automation
An Agent Automation can be created in the following ways:
- Manually from the AI Command Center UI
- Programmatically via APIs
Manually Initiating an Agent Automation
- Navigate to Agents → Agents
- Select the required Agent
- Click Run
Initiating an Agent Automation via API
External applications can initiate Agent Automations using IntelliBuddies Client APIs.
This approach is typically used for system-to-system integration scenarios.
Activity Category License Enforcement
Agent execution is governed by Activity Category–level licensing.
Each activity belongs to an Activity Category (for example, Primitive, Integration, RPA, AI).
An Agent Automation can execute an activity only if the associated subscription includes a valid license for that category.
How License Enforcement Works
- License validation is performed at runtime, just before an activity executes.
- The system verifies whether the subscription includes the required activity category.
- Validation applies uniformly across all Agent Automations
Enforcement Behavior
- If the required activity category license is available, execution proceeds normally.
- If the license is missing:
- The activity execution is blocked.
- The Agent Automation transitions to a Faulted or Suspended state (based on the configured error strategy).
- A clear error message is recorded in:
- Activity logs
- Automation logs
- Error Details view
User Feedback
When execution is blocked due to licensing:
- The system clearly identifies the missing activity category
- Explains why execution was restricted
- Advises users to contact an administrator or licensing manager
This ensures transparent and predictable runtime behavior aligned with subscription entitlements.
Suspending an Agent Automation
Suspending an Agent Automation pauses execution and suspends all associated active tasks.
Suspend a Single Automation
- In the Automations page
- Select the Agent
- Click Suspend
Suspend Multiple Automations
- Select multiple automations using checkboxes
- Click Suspend in the toolbar
Suspending an automation also suspends all associated human and buddy tasks.
Resuming a Suspended Agent Automation
A suspended automation can be resumed manually or programmatically.
To resume from the UI:
- Navigate to Reports → Automations
- Select the suspended automation
- Click Resume
Cancelling an Agent Automation
Cancelling permanently stops a running or suspended automation.
- Select the agent automation and click Cancel
- For bulk cancellation, select multiple agent automations and click Cancel in the toolbar
Changing the Execution Flow of an Agent Automation
Authorized users can move execution forward or backward to a specific activity.
Steps:
- Open the automation
- Navigate to the Process Monitor tab
- Select the target activity
- Click Move To
- Requires the Initiate Agent Automation permission
- Not available in embedded Process Monitor views
- Rolling back retains previously entered form data
- Rolling forward may cause data loss if activities are skipped
- In parallel flows, other branches may still need completion
Agent Automation Audit
The Automations page provides access to detailed audit logs for every automation.
These logs record user actions and system events, enabling traceability, compliance, and operational troubleshooting.
You can open the audit logs by selecting an automation and choosing Audit.
Agent Automation Audit Logs
The Audit Log captures all significant actions performed on an automation by users or the system.
| Event | Description |
|---|---|
| Automation Created | A new Agent Automation was created. |
| Automation Viewed | A user opened the automation details page. |
| Automation Suspended | A running automation was suspended. |
| Automation Resumed | A previously suspended automation was resumed. |
| Automation Cancelled | The automation was cancelled. |
| Moved to Target Activity | Execution was manually redirected to a specific activity. |
| Session Count Updated | The session or concurrency count was modified. |
| Automation Data Updated | One or more automation-level data values were edited. |
| Model Data Updated | One or more model data values were updated during execution. |
Purpose of Audit Logs
Agent Automation Audit Logs provide:
- Operational traceability – Track how an automation was handled
- User accountability – Identify who performed specific actions
- Compliance support – Maintain a complete activity trail
- Troubleshooting visibility – Understand how an automation reached its current state
Audit logs are immutable and form the authoritative history of each automation.
Overriding Maximum Allowed Sessions
Each Agent has a configured Maximum Allowed Sessions value.
If this limit is exceeded, the automation is automatically suspended.
Authorized users can override this value:
- Select the automation and click Edit Max Session Count
- Or open the automation and use the toolbar option
Agent Automation Details Page
The Agent Automation details page contains several tabs that provide visibility into execution, data, and diagnostics.
General Information
| Information | Description |
|---|---|
| Agent Name | Shows the name of the Agent executed by this automation. |
| Agent Version | Indicates the version of the Agent executed. |
| Started Date | Displays the start date and time of the automation. |
| Completed Date | Shows the completion date and time of the automation. |
| Execution Time | Indicates the total duration of the automation. |
| Automation Name | Shows the specific name assigned to the automation. |
| Automation ID | Indicates the unique identifier assigned to the automation. |
Agent Monitor
Provides a visual representation of the execution flow, showing:
- Currently executing activities
- Completed activities
- Faulted activities
- Execution duration per activity
You can double-click an activity to view execution details.
Viewing Task Data
For task-based activities (eForms), use View Data to see submitted form data after task completion.
Activities
Displays all activities executed as part of the automation.
Columns include:
Workflow, Display Name, Session, Status, Started At, Ended On, Duration
Activity Status
| Status | Description |
|---|---|
| Active | Activity is currently executing |
| Passed | Activity completed successfully |
| Faulted | Activity failed during execution |
Viewing Error Details
Double-click a faulted activity to view:
- Error message
- Exception type
- Source
- Stack trace
- Additional diagnostic data
Human Tasks
Displays all human tasks created by the agent automation.
- Completed task data can be viewed using View Data
- Available only for completed task activities
Buddy Tasks
Displays all buddy (RPA) tasks created by the agent.
- Execution logs can be viewed by double-clicking a task
- Includes both active and completed buddy executions
Agent Data
Displays the latest runtime values of Model Data and Form Data.
Editing Agent Data
To edit data:
- User must have Administrator role
- or Update Agent Automation Data permission
Double-click a leaf property and update the value
Track Changes
A history icon appears when data is modified.
- Selecting the icon displays the full change history.
Logs
Displays all logs generated during Agent execution, including:
- System logs
- Activity execution logs
- Error and diagnostic messages
Notifications
Displays Agent Automation–specific notifications generated during execution.
- Agent initiators and users with Manage All Agent Automations permission can view all notifications
- Other users can view only notifications addressed to them
Summary
The Agent Automation page is the primary operational interface for monitoring, controlling, and diagnosing Agent execution.
With the introduction of Activity Category–level licensing, this page plays a critical role in surfacing runtime enforcement behavior, errors, and execution state in a transparent and actionable manner.