Agents

In Riverbed IQ Ops, agents are how intelligence runs inside automation, not only in conversation. Agents connect catalog definitions to runbooks An automated workflow that executes a series of steps or tasks in response to a triggered event, such as the detection of anomalous behavior generating an incident, a lifecycle event, or a manually executed runbook. so repeatable work does not depend on an operator opening Chat.

For the broader story, see AI-Assisted Operations in Riverbed IQ Ops. For how to move around the left menu, see Moving Around Riverbed IQ Ops.

Agents in runbooks

The Agentic AI runbook node runs a selected catalog agent as part of a runbook An automated workflow that executes a series of steps or tasks in response to a triggered event, such as the detection of anomalous behavior generating an incident, a lifecycle event, or a manually executed runbook. branch. Use it when a flow should call an agent with runbook context (variables, upstream outputs A document containing data sets generated by the execution of a runbook, including output of queries and reports from point products, as well as output of analysis or other runbook nodes., and triggers) rather than a one-off chat message. See Agentic AI.

Agent catalog and editor

The Agent Catalog page is where you browse existing agents, filter by mode, state, owner, and category, and open an agent to edit. Open it from Automation > AI Agents.

Click + New Agent to open the create flow.
Open an existing card to edit agent details.
Use Back to Catalog in the editor header to return to the catalog list.
Use Create Agent Runbook to open on-demand runbook creation with an Agentic AI node already set to the current agent.
Use View Schedules to open the Runbook Scheduling page and manage schedules for on-demand runbooks.

Click + New Agent or open an existing card to work in the agent editor. See Creating and Editing Agents.

Chat with catalog agents

On the Chat page, use the agent chooser in the page header to start a session with a published catalog agent instead of the default assistant, Riverbed Q. This is useful when you want an agent specialized for a domain (for example, troubleshooting a specific application) without wiring a runbook first. The page title shows Chat Session with followed by the selected agent's name.

Scheduling agents through runbooks

Note: You do not schedule an agent catalog entry by itself. You schedule an on-demand runbook that includes an Agentic AI node for that agent.

To run an agent on a timetable:

Publish the agent in the Agent Catalog.
Create an on-demand runbook that contains an Agentic AI node for that agent. From the agent editor, click Create Agent Runbook to start with that node already configured.
Open Automation > Scheduling (or click View Schedules in the agent editor) and schedule that on-demand runbook. See Runbook Scheduling page.

Catalog entries of type Command can supply explicit operational invocations that agents and Chat use. Pair commands with on-demand runbooks and scheduling when you need recurring automation. See Skills and Commands Berkeley Packet Filter. A syntax for selecting which network packets are captured or harvested based on protocol, address, port, and other packet attributes..

The Agent Catalog can show modes such as Scheduled and Sentinel Mode on agent cards (filters use Scheduled and Sentinel). Those modes describe how the agent is intended to be used. They are not a separate scheduler for the agent entry. When an agent is published, catalog filters use Published and cards show the status badge Active.

What agents are not

Agents in this documentation refer to Riverbed IQ Ops catalog and runbook agents. They are not the same as the separate Unified Agent experience in Network Observability (the Aternity agent platform entry in Apps). If you need endpoint or NPM agent deployment, use that product documentation instead.

Build a useful agent practice

Publish proven instructions as skills and operational invocations as commands Berkeley Packet Filter. A syntax for selecting which network packets are captured or harvested based on protocol, address, port, and other packet attributes. before wiring runbooks.
Use the Agentic AI node where runbook data should shape the agent request.
Use Chat with Riverbed Q (Q) to prototype logic, then codify results in the Skill Catalog.
Schedule the on-demand runbook that calls the agent when you need unattended, recurring execution.

AI-Generation disclaimer

When using Generative AI technologies to create and generate content, the target audience must be aware that the content is AI generated. We recommend adding a disclaimer to the output of the content generated using the GenAI node. For example, appending "AI-Generated content".