Editing and Duplicating Agents in Sitecore Agentic Studio

In my previous article, I explored the core concepts of Sitecore Agentic Studio and explained how AI agents can help automate content creation and marketing workflows. While Sitecore provides several prebuilt agents to get started, you don't always need to create a new agent from scratch when your requirements change.

Instead, you can edit an existing agent to improve its behavior or duplicate an agent to create a customized version for a different use case.

Understanding when to edit, duplicate, or create a new agent can save time, promote consistency, and help you build reusable AI solutions.

In this article, we'll explore both approaches with practical examples and walk through the steps involved.

Why Modify an Existing Agent?

As organizations begin using AI more frequently, their requirements naturally evolve.

For example:

  • Your marketing team wants a different writing tone.
  • Developers need more technical documentation.
  • Product teams require content for a different audience.
  • You discover a better prompt that produces higher-quality results.

Rather than creating multiple agents from scratch, Sitecore Agentic Studio allows you to reuse existing agents by editing or duplicating them.

Edit vs Duplicate – Which One Should You Choose?

Although both options allow customization, they serve different purposes.

Edit an Agent

Duplicate an Agent

Updates the existing agent

Creates a new copy

Changes affect everyone using that agent

Original agent remains unchanged

Best for improvements and bug fixes

Best for creating variations

Maintains the same identity

Creates a completely separate agent

A simple rule to remember is:

  • Edit when you want to improve the existing agent.
  • Duplicate when you need another agent with similar functionality.

Scenario

Imagine your team already has a Technical Blog Generator agent.

The current agent generates articles using a neutral technical writing style.

Now different teams have new requirements:

Developer Relations Team

They want:

  • Detailed explanations
  • Architecture diagrams
  • Code samples
  • Best practices

Marketing Team

They need:

  • Beginner-friendly language
  • Customer-focused messaging
  • Call-to-actions
  • SEO optimization

Creating both versions from scratch would mean rewriting prompts, inputs, and workflows.

Instead, you can duplicate the existing agent and customize only the parts that differ.

Editing an Existing Agent

Editing allows you to improve an existing agent without changing its identity.

Typical reasons include:

  • Improving prompts
  • Adding instructions
  • Fixing incorrect outputs
  • Updating workflows
  • Improving formatting
  • Adding new input parameters

Because you're modifying the original agent, all users of that agent will benefit from the updates.

How to Edit an Agent

Step 1

Open Agentic Studio and navigate to the Agents section.

Locate the agent you want to modify.

Step 2

Select the agent to open its configuration.

The agent editor displays several sections, including:

  • Overview
  • Parameters
  • Context
  • Workflow
  • Schemas
  • HTML Templates
  • Inputs and Outputs

Step 3

Update the required configuration.

For example, improve the Instructions like

Before

Generate a technical article.

After

Generate a technical article for experienced Sitecore developers.

Include:

- Introduction

- Architecture

- Code examples

- Best practices

- Common pitfalls

- Summary

Use Markdown formatting.

This additional context helps the AI generate more consistent and useful content.

Or

Add new steps in the workflow

 

Step 4

Save the changes.

Run the agent using sample inputs and review the generated output.

Continue refining until the responses consistently meet your expectations.

Example: Improving a Blog Generator

Suppose your original prompt generates articles that are too generic.

Instead of creating a new agent, simply update the instructions.

Before

Write a blog about Sitecore Search.

After

Write a detailed technical blog about Sitecore Search.

Audience:

Intermediate Sitecore Developers

Include:

- Overview

- Architecture

- Configuration

- Practical Example

- Best Practices

- Conclusion

Keep the tone technical.

The updated instructions produce richer and more structured articles while preserving the same reusable agent.

Duplicating an Agent

Duplicating creates an entirely new agent using an existing one as a starting point.

The duplicate contains:

  • Instructions
  • Inputs
  • Workflow
  • Schemas
  • HTML templates

Once duplicated, the new agent can be modified independently.

Changes made to the duplicate do not affect the original agent.

When Should You Duplicate?

Duplicating is useful when:

  • Multiple departments need similar agents.
  • Different writing styles are required.
  • Different workflows are needed.
  • You want to experiment without affecting production.
  • You need customer-specific agents.

Example: Creating Multiple Blog Writers

Suppose your company already has this agent:

Technical Blog Generator

You can duplicate it to create:

Developer Blog Generator

Focuses on:

  • Code samples
  • APIs
  • Architecture
  • Troubleshooting

Marketing Blog Generator

Focuses on:

  • Customer benefits
  • SEO
  • Simple language
  • Call-to-actions

Documentation Generator

Focuses on:

  • Step-by-step guides
  • Product documentation
  • Release notes

Each agent shares the same foundation while serving a different audience.

How to Duplicate an Agent

Step 1

Navigate to Agentic Studio → Agents and Locate the existing agent.

Open the More Actions (⋮) menu and Select Duplicate.

Provide a new name.

Example:

Blog Generator – Developer Specific or Blog Generator - Marketers

Update the description to clearly identify its purpose.

 

 

 

Audience:

Business users

Tone:

Friendly and engaging

Adjust Inputs if required.

For example, add:

  • Target Audience
  • Keywords
  • Campaign Goal

Save the duplicated agent.

Test it with several prompts to ensure it produces the desired output.

Best Practices

When working with agents, consider the following recommendations:

  • Keep each agent focused on a single purpose.
  • Use descriptive names that clearly identify the agent's role.
  • Test agents with different types of input before sharing them with your team.
  • Duplicate an existing agent whenever only small changes are required.
  • Edit an agent only when the changes should benefit all users.
  • Review and refine prompts regularly as business requirements evolve.

Common Mistakes

Avoid these common pitfalls:

Editing when you should duplicate - If multiple teams rely on the same agent, editing it may unintentionally change their workflows.

Creating too many similar agents - If the only difference is a minor prompt adjustment, consider whether editing or parameterizing the existing agent is sufficient.

Poor naming - Names like "Agent1" or "Copy of Blog Generator" make it difficult to identify the right agent later.

Instead, use descriptive names such as:

  • Technical Blog Generator
  • Marketing Blog Generator
  • Product Documentation Assistant
  • SEO Content Generator

Edit, Duplicate, or Create?

The following guidelines can help you decide:

Requirement

Recommended Action

Improve an existing prompt

Edit

Fix incorrect behavior

Edit

Add new workflow steps

Edit

Create another version for a different audience

Duplicate

Build an experimental version

Duplicate

Solve a completely new business problem

Create a new agent

Conclusion

One of the strengths of Sitecore Agentic Studio is that it encourages reuse rather than duplication of effort. By understanding when to edit an existing agent and when to duplicate one, you can build a well-organized library of AI assistants that evolve alongside your business needs.

Editing helps improve shared agents over time, while duplicating enables teams to create specialized versions without impacting existing workflows. Choosing the right approach not only saves development time but also promotes consistency, maintainability, and collaboration across your organization.

As your collection of agents grows, adopting these practices will make it much easier to manage, scale, and maintain AI-powered solutions in Sitecore Agentic Studio.

 

Comments

Post a Comment

Popular posts from this blog

Migrating from Sitecore JSS SDK to Content SDK – Best Practices and Pitfalls

Image
The Sitecore JSS SDK has been a staple for building modern headless websites with XM Cloud and Sitecore XP. It offers deep integration with the Sitecore Layout Service, letting frontend developers render content directly based on Sitecore-managed components. Now, with the introduction of the Sitecore Content SDK, teams have a lightweight, REST-first way to fetch clean JSON data, leaving rendering entirely to the frontend. For projects focused on performance, composable architectures, or custom rendering logic, the Content SDK can be a compelling choice. If you’re considering a switch from JSS SDK to Content SDK, this guide walks you through how to migrate, best practices, and the pitfalls to avoid. Why Migrate to Content SDK? Lightweight: No dependency on Layout Service rendering. Faster payloads: Returns only the requested fields, reducing over-fetching. Flexible rendering: Total control of frontend presentation. Better for static generation: Idea...
Read more

Beginner’s Guide: Step-by-Step Local Installation for Sitecore XM Cloud

Image
Many times, we want to accomplish a task but aren’t sure where to start—the same applies to Sitecore XM Cloud local setup and installation. In this guide, I’ll Walk you through the process of setting up a Sitecore XM Cloud environment locally, step by step. Sitecore XM Cloud is the cloud-native, headless version of Sitecore Experience Manager (XM), designed for modern, omnichannel digital experiences. It separates content management (authoring) from content delivery, enabling fast, API-driven, and scalable architectures. Developers can deliver content to web, mobile, and other channels via GraphQL APIs and Experience Edge, while content authors manage items in a cloud-first environment. Even though XM Cloud is cloud-first, a local environment is essential for: Experimenting safely with content, templates, and workflows. Testing headless frontends like Angular, React, or Next.js. Debugging GraphQL queries and APIs efficiently. Validating scheduled tasks...
Read more

Working Locally with Sitecore XM Cloud: Connecting, Managing, and Serializing Items

Image
In the previous blog , we walked through the step-by-step local installation of Sitecore XM Cloud. In this post, we’ll take the next step by exploring how to connect cloud and local environments setup by: Configuring item serialization Connecting to an XM Cloud environment Connecting to your local environment Pull serialized items from XM Cloud into your local repository Push serialized items into the local environment This workflow is essential because a local XM Cloud instance is a vanilla instance with no items. To work effectively, you’ll need to bring down items from your cloud environment. With Sitecore Serialization, you can sync items between cloud and local setups, ensuring your local development mirrors what exists in the cloud. Configure item Serialization The local installation includes the file at the below path for rendering host serialization authoring/items/next-starter.module.json For local development, you need a new ...
Read more