KentiCopilot

Building with Xperience by Kentico means holding a lot of context – platform APIs, project conventions, and your content model. KentiCopilot gives that context to the AI coding assistant you already use.

Meet KentiCopilot

KentiCopilot is a suite of AI development tools that enables agentic development for Xperience by Kentico projects. It combines MCP (Model Context Protocol) servers, curated skills and workflows, and best practices to help developers work more effectively with AI coding assistants when building with Xperience by Kentico.

Instead of acting as a standalone AI agent, KentiCopilot tools bridge your preferred AI development tools with the Xperience by Kentico platform. They work with any AI coding assistant that supports the Model Context Protocol (MCP) and agent skills.

The following graph shows the available KentiCopilot tools and resources.

Add KentiCopilot into your project

Installing the tools is the first step. An agent also needs to know how your project is laid out, which conventions it follows, and how to verify its own changes. Without that context, every session starts over and the results vary.

Install the KentiCopilot plugins

KentiCopilot skills are distributed as agent plugins, and the KentiCopilot repository serves as the plugin marketplace. Each plugin installs separately, so you can add only the ones your work needs. The usage guide in the repository lists the available plugins and the detailed installation steps.

Add the marketplace to your coding assistant, then install a plugin by name. For example, in Claude Code:

/plugin marketplace add Kentico/xperience-by-kentico-kenticopilot
/plugin install kentico-web-development@xperience-by-kentico-kenticopilot

The commands differ between AI tools. See the usage guide for the equivalent steps in yours.

Check your project’s readiness

The agentify skill in the web development plugin audits an existing project against Kentico’s agentic-development practices. The skill asks which AI tool you work with, evaluates the project, and writes a readiness report that names each gap. Nothing in your repository changes until you approve the recommendations.

To run the audit, you need an Xperience by Kentico project and the web development plugin. Then ask your assistant to run the agentify skill against the project in your workspace.

The audit evaluates your project in a few areas:

• Agent instructions
• Design and coding standards guidance
• Documentation and Management MCP server configuration

With your approval, the skill then creates the missing files and places them where your AI tool discovers them.

A project that passes the audit gives every new session the same starting context, which makes the results of the other curated skills and workflows more consistent.

With the project prepared, see Work effectively with KentiCopilot for the habits that get the most out of the tools.

Curated skills and workflows

Effective AI-assisted development requires more than just access to tools. The KentiCopilot tools include a growing library of skills and plugins specifically optimized for Xperience by Kentico development scenarios.

Pre-built skills and orchestrated workflows help you with common development tasks:

• Prototyping your content model, including content types, taxonomies, and Page Builder components.
• Developing Page Builder widgets, sections, and page templates.
• Building custom modules and integration scaffolding built on top of real-life examples.
• Analyzing of content types and their relationships and programmatic content types creation.

AI-assisted web development

The web development plugin covers a design-led implementation end to end: preparing the project for agentic development, modeling content from a design, building Page Builder widgets, sections, and page templates, writing content retrieval code, and comparing the finished implementation against the original design.

Before generating any code, these skills study the components your project already has, mirror your repository’s conventions, and verify uncertain APIs against the current Xperience documentation.

Development capabilities:

• Generated artifacts: Content types and reusable field schemas derived from a design. Page Builder widgets, sections, and page templates with their views, properties classes, and localized resources. Content retrieval code for the models you build. If your repo already has components, the output mirrors your existing filesystem layout and coding patterns.
• Context files: Shipped alongside the skills, they instruct the agent to look for specific resources and information, such as core concepts and APIs, respective documentation pages, or example components to help the agent identify implementation patterns.
• Best practices for the components you build, with examples and customization ideas.

Check out a video series where KentiCopilot powers a full widget build. See how AI automates content modeling and creates a ready-to-use Page Builder widget in Build Smarter: From Content Model to Widget with KentiCopilot.

KX13 upgrade hands-on assistance

KentiCopilot includes specialized tooling to help you migrate legacy Kentico Xperience 13 (KX13) projects to Xperience by Kentico. The KX13 migration plugin covers the full upgrade journey in a single plugin, organized into groups of skills:

• Content-model audit – Pairs an AI skill with a bundled .NET CLI that snapshots a KX13 database into structured JSON and a Markdown report. The output is the canonical input for the migration plan.
• Content migration – Turns the audit into a migration plan, generates the Kentico Migration Tool appsettings.json and the C# extensions for Migration.Tool.Extensions, executes the migration tool, and evaluates the result. The eval skill produces an HTML report that classifies each finding and routes back to the specific sibling skill to re-run.
• Codebase migration – Sets up the XbyK project foundation, migrates page controllers and views to the new content retrieval API, migrates Page Builder widgets and shared components, and reconciles visual differences between original and migrated pages with Playwright.

Upgrade capabilities:

• Source content audit: Snapshots the KX13 content model into structured JSON, giving the rest of the skills a deterministic input.
• Migration plan generation: Decides content-type strategy and migration tool configuration from the audit – content hub conversions, reusable field schemas, linked-page handling, widget transforms.
• Automated code analysis: Scans the KX13 codebase to identify APIs, patterns, and dependencies that require updates.
• Pattern Transformation: Converts legacy code patterns to XbyK equivalents with AI assistance.
• Iterative loop: Treats configure → codegen → run → eval as one loop. The eval skill names the specific sibling skill to re-run for each finding.
• Incremental migration: Supports gradual migration in manageable steps – feature-by-feature, page-by-page.

Project lifecycle maintenance

Once a project is in production, the project lifecycle plugin helps you keep it current and deploy changes in a controlled way:

• update-xperience – Determines your current and target Xperience versions, reviews every Changelog entry in between together with the update guides they link to, and then follows the official update procedure.
• cd-repository-configure – Builds deployment-scoped repository.config filters from the pull requests or commit range you want to deploy, excluding Xperience version-only changes, and then exports and verifies the serialized deployment content.

Digital experience extensions

The digital experience plugin enables coding assistants to develop custom components for automation processes. Its skills implement and register custom automation steps and triggers, which marketers can use and configure in the Automation Builder.

MCP servers

Model Context Protocol (MCP) servers are the foundation of KentiCopilot’s capabilities. These servers expose Xperience by Kentico functionality directly to AI assistants and enable them to interact with your projects, access documentation, and execute Kentico-specific operations in real time.

MCP servers use a standardized protocol that allows AI assistants to discover available capabilities, understand their parameters, and execute operations safely within your development environment.

Documentation MCP server

Provides access to Xperience by Kentico documentation and code examples. Serves as a reliable source of context for your agentic workflows. See Documentation MCP server to learn more.

Content Modeling MCP server

Provides content modeling assistance with validation and best practice checks. See Content Modeling MCP Server to learn more.

Management MCP server

Provides integration with the Xperience by Kentico management API to allow querying and editing of content, configuration, and other objects stored in the system’s database. This includes the project’s content model (content types and reusable field schemas), content items, pages (including Page Builder content and configuration), channels, languages and more.

See the Management MCP server documentation for more information and instructions.

Open-source MCP servers and libraries

We’ve built and open-sourced the following tools to help you deliver your projects more easily:

• Virtual Inbox MCP server – An in-app hosted virtual inbox for email channel development, testing, and AI agent email access. You can use it when you need AI agents to validate email content and rendering after code changes, drive end-to-end testing of visitor journeys, such as automations and membership experiences, or run demo environments that need email “delivery” without a real SMTP inbox.
• Component Registry MCP server – This custom module lets administrators see all registered custom components in an Xperience by Kentico app, like Page Builder widgets. It also shows which channels and web pages use each component through an admin interface. Read the blog post to learn how and why we built this module.
• Annotated Kentico icons library – This is a ready-to-use, annotated list of all icons used in Xperience by Kentico Page, Email, and Form Builder components. It helps AI agents avoid making up icons when building Xperience components. To learn more about why we created this library, check out the article Smarter Components: Using Icon Metadata to Guide AI in Kentico Projects.
• Admin UI Design Components – A modified reference library for the React components that power the Xperience by Kentico administration UI, which gives AI agents component design context when generating admin UI extensions.

Custom MCP servers

In addition to the official Kentico MCP servers that provide out-of-the-box functionality, you can create your own tooling, such as custom MCP servers and other toolsets:

• Apply custom business logic and domain- or agency-specific operations.
• Integrate with your organization’s internal tools and systems.
• Provide access to project-specific shortcuts and automation workflows.
• Promote agency-specific or team-specific conventions and patterns.

The official Kentico MCP servers are compatible with third-party MCP servers, so you can extend the current functionality with your own servers or integrate other tools into your development workflow.

Learn AI-assisted development fundamentals

KentiCopilot works best when you are comfortable directing an AI coding assistant. If you are new to AI-assisted development, start with the general fundamentals. The same habits apply to every assistant: plan before you generate code, provide specific context instead of vague prompts, and review everything the assistant produces.

The following free resources assume little or no prior AI knowledge:

• Anthropic Academy – Free self-paced courses, ranging from everyday AI assistant usage to developer topics such as MCP and agent skills.
• AI Fluency: Framework & Foundations – Teaches the four core competencies of working with AI: Delegation, Description, Discernment, and Diligence. Suitable for any role.
• Claude Code best practices – Explains how to explore and plan before coding, provide specific context in prompts, and delegate investigation to subagents. The principles transfer to other agentic coding tools.
• How an agent remembers your project – Shows how persistent instruction files, such as AGENTS.md, give an assistant project context across sessions.
• GitHub Spec Kit (advanced) – A toolkit for spec-driven development. You describe what to build in a specification, and the assistant derives the plan, tasks, and code from it. Works with GitHub Copilot, Claude Code, and other agents.

Kentico-specific materials build on those fundamentals. You can find them in the Xperience by Kentico documentation, in the developer guides, or in blog posts on the Community Portal:

• Step-by-step guidance for installing the plugins and invoking their skills.
• Conceptual guides explaining how Xperience by Kentico MCP servers work and how you can put them to work.
• Migration guides for adopting AI-assisted development in existing projects.
• Query and data retrieval patterns based on Kentico documentation and API examples.

Once you know the fundamentals, the KentiCopilot curated skills and workflows apply them to Xperience by Kentico scenarios, so you do not need to design your own prompts and workflows from scratch.

AI tool compatibility

KentiCopilot is built on open standards, primarily the Model Context Protocol (MCP) and agent skills. This means you’re not locked into a specific AI tool, and no specific tool or editor is required. See System requirements for more information.

As new AI development assistants emerge, KentiCopilot’s MCP and skill-based architecture ensures compatibility without major changes to your workflow.

Some of the KentiCopilot curated skills and workflows use plugin architecture, which allows quick installation specifically in Claude Code and GitHub Copilot environments.

KentiCopilot guides and videos

Working with KentiCopilot

• Work effectively with KentiCopilot CONCEPT – Learn the habits that get better results from any AI tool: explore before coding, delegate investigation to subagents, give specific context, and verify the result against the running site.

Documentation access

• Documentation MCP server – Learn how to install the Documentation MCP server to provide your AI assistant with real-time access to Kentico documentation through semantic search and content retrieval. The MCP server is compatible with any IDE or AI client that supports MCP.

Content type management

• Analyze content type relationships HOW-TO – Use your agentic coding assistant to generate a concise summary of content types and their relationships.

Widget implementation

• Rapidly implement widgets with KentiCopilot HOW-TO – Watch a five-minute demo showing widget implementation from start to finish, then use the detailed skills and requirements templates from the KentiCopilot repository to build your own widgets with AI assistance.
• Implement product display patterns with KentiCopilot CONCEPT – Examine key considerations for rapidly creating product display widgets with KentiCopilot’s plugins and skills.

Remodeling during upgrades

• Speed up remodeling with AI CONCEPT – See how agentic AI coding assistants can accelerate the process of remodeling content types and field mappings during your upgrade. Learn how to provide the best context and instructions for the highest quality results.

AI-powered content modeling

• Content Modeling MCP Server CONCEPT – Learn how to use the Content Modeling MCP server to speed up content model creation with a structured workflow and built-in validation against Kentico best practices. The MCP server is compatible with any IDE or AI client that supports MCP.

• Prototype your model with Content modeling MCP server CONCEPT – See how you can use the Content Modeling MCP server together with the Management MCP server to rapidly prototype a content model for your project. Learn how to speed up model creation, iterate faster, and keep everything consistent across your solution.

Community case studies

• AI assisted improvements to the Kentico Community Portal – Learn how GitHub Copilot was used to accelerate UI improvements, search redesign, and Q&A tagging on a real in-production Xperience by Kentico project.

• Can AI really update my Xperience by Kentico project? – Explore an AI-driven workflow for automating Xperience by Kentico hotfix and refresh updates, including package changes, database migrations, and breaking change resolution.

• AI-Powered Refactoring: Updating Xperience by Kentico Logging with a Single Prompt – See how a single GitHub Copilot prompt, backed by the Documentation MCP server, migrated the entire Kentico Community Portal from IEventLogService to ILogger<T>.

• Deliver More: AI Assisted Improvements to Q&A Discussions – Discover how AI expanded the scope of Community Portal improvements, including discussion notifications, timezone customization, and reaction voting, without increasing development time.

• Smarter Components: Using Icon Metadata to Guide AI in Kentico Projects – Learn how AI-generated descriptions added to Kentico’s icon class help AI agents select the right icon when generating Page Builder component registrations.

KentiCopilot FAQs

Is KentiCopilot an AI agent?

No, KentiCopilot is not an AI agent itself. It’s a suite of tools that enables developers by enhancing 3rd-party AI assistants. The tools developed under the KentiCopilot umbrella provide AI assistants and agentic coding tools with specialized knowledge and capabilities for Xperience by Kentico development. Think of the KentiCopilot tooling as giving your AI assistant expert-level knowledge of Kentico, but not turning it into an autonomous stand-in for developers.

Is KentiCopilot tooling for C#/.NET only, or does it include also JavaScript/React?

The toolkit recognizes that Xperience by Kentico projects often involve both server-side and client-side code. Tools built under the KentiCopilot umbrella support both back-end and front-end development:

• C#/.NET back-end development of ASP.NET Core applications, custom modules, integrations.
• JavaScript/TypeScript front-end development, including Xperience by Kentico React components.