Invoke the Network Security Agent
Focus
Focus
Strata Cloud Manager

Invoke the Network Security Agent

Table of Contents

Invoke the Network Security Agent

Submit natural language queries and use contextual UI prompts to start agent workflows in Strata Cloud Manager.
Where Can I Use This?What Do I Need?
  • Strata Cloud Manager
  • Strata Cloud Manager Pro
  • This feature is currently in Beta and available on request. Contact your account team to enable the feature.
  • Superuser role (for agent administration, onboarding, and managing permissions)
  • Network Administrator or Security Administrator role (for agent interaction)
This feature leverages a large language model or generative AI, and may contain errors. Double-check for accuracy before applying changes.
You can invoke the Network Security Agent reactively through natural language queries or contextual UI clicks. In both cases, the agent generates a Target Plan for your review before any actions are executed. Use the Expert Agent alias to route your query directly to a specific domain, or use @agent to let the Planner determine the best Expert for your request.

Invoke Using Natural Language

  1. Open Strata Copilot in Strata Cloud Manager.
  2. Select an agent from the @Agent drop-down, or type @agent followed by your request in the query field. For example:
    • @agent/troubleshooting Check tunnel status for branch-office-3
    • @agent/threat Am I protected from CVE-2026-1234?
  3. Use auto-complete suggestions as you type. Typing @D suggests @Deployment, typing @T suggests @Threat, and so on.
  4. (Optional) Select agents from the Agents button in the query context ribbon below the input field. Select one or more agents, then enter your query.

Invoke Using Contextual UI Prompts

  1. Open a Strata Cloud Manager page that supports contextual agent invocation (for example, the policy configuration page).
  2. Select the Agent Assist button or contextual prompt embedded on the page.
  3. The agent opens a conversation pre-loaded with the context of that page, including any active filters or session context you applied.

Review and Execute an Agent Plan

After you or another user invokes the agent, it generates a Target Plan—a proposed sequence of steps to achieve the stated objective. You must review and approve the Target Plan before execution begins.
  1. Review the Target Plan displayed in the conversation. The plan includes:
    • A human-readable description of the objective.
    • An ordered list of steps the agent proposes to execute.
  2. (Optional) Modify the plan before execution:
    • Adjust parameters using natural language (for example, "Change the cleanup threshold from 90 days to 365 days").
    • Natural language modifications are validated to confirm they are relevant to network security and do not contain malicious intent before being applied to the plan.
  3. Execute to approve the plan.
  4. Monitor progress in the conversation view:
    • Basic Conversation: High-level progress indicators and outcome summaries.
    • Detailed Conversation: Reasoning steps, data sources, and detailed tracing.
  5. When a write task requires execution in Supervised mode, the plan pauses for you to approve or reject each write action.
  6. After execution completes, review the Execution Summary that summarizes objectives, outcomes, recommended next steps, and linked write references.