Introduction
Generative AI is everywhere, and Adobe is no exception. From content creation to software development, Adobe is integrating AI across its ecosystem to help users work smarter and faster.
In this blog series, we’ll simplify Adobe’s Gen AI ecosystem and explore the tools that can make a real difference in your day-to-day work. In Part 1, we’ll look at Adobe Skills and how they help AI coding assistants better understand Adobe technologies, especially Adobe Experience Manager (AEM).
What Are Adobe Skills?
AI coding assistants are great at generating code, but they often lack deep product-specific knowledge. This can result in suggestions that don’t always align with Adobe’s recommended implementation patterns.
Adobe Skills (officially known as Agent Skills) solve this problem by providing AI coding assistants with Adobe-specific knowledge and workflows. Instead of relying only on general AI knowledge, they can perform tasks using Adobe’s best practices and recommended approaches.
Installing Adobe Skills
If you want to install all available Adobe Skills, run:
npx skills add adobe/skills –all
If you’re only working with AEM as a Cloud Service, install only the AEM Skills:
npx skills add https://github.com/adobe/skills/tree/main/plugins/aem/cloud-service
During installation, you’ll be prompted to select the skills you want to install.
Press the Space key to select or deselect individual skills and select the agentic IDE/CLI you’re using, for example, Cursor, Claude, etc.
What Happens Behind the Scenes?
Installing Adobe Skills does much more than simply add a few prompts to your AI coding assistant. It automatically prepares your local development environment so the assistant can understand your AEM project better.
- Installs the selected Adobe Skills.
- Creates symbolic links so your AI coding assistant can discover and use the installed skills.
- Generates or updates the AGENTS.md file, giving the AI assistant project-specific context and instructions.
- Registers slash commands that allow individual skills to be invoked directly when needed.
- Configures the skills so they work seamlessly with supported AI coding assistants.
After the installation completes, the skills are immediately available and can be triggered automatically based on your prompts, without requiring any additional configuration.
Using Adobe Skills
Once the installation is complete, you can start using natural language prompts with your AI coding assistant.
For example:
Scan my AEM project and report any code quality issues.
Instead of performing a generic review, the AI automatically invokes the code-assessment Skill to analyse your project using Adobe’s recommended AEM development practices.
Common Use Cases
Adobe Skills are designed around real AEM development workflows. Based on your prompt, your AI coding assistant automatically invokes the most appropriate skill.
| Use Case | Example Prompt | Skill Invoked |
|---|---|---|
| Create AGENTS.md for my project | Generate AGENTS.md for my AEM project. | ensure-agents-md |
| Develop a new AEM Component | Create a Hero Banner component with a dialog, Sling Model, HTL, Clientlibs, and unit tests. |
create-component |
| Migrate an existing AEM project | Migrate my AEM 6.5 scheduler implementation to AEM as a Cloud Service. |
migration |
| Configure Dispatcher | Help me configure Dispatcher caching for this URL and optimize cache rules. |
dispatcher |
| Create or Debug Workflows | Create an approval workflow with email notifications, or help debug a failed workflow. |
aem-workflow |
The AI identifies the appropriate Adobe Skill based on your request, so you don’t need to remember or invoke the skill name manually.
Conclusion
Adobe Skills are one of the easiest ways to make your AI coding assistant more effective for Adobe development. With just a few commands, you can equip your assistant with Adobe-specific knowledge and receive recommendations that are far more relevant than generic AI responses.
In Part 2, we’ll explore Local MCP Servers for AEM and see how they provide AI coding assistants with live project context, making them even more powerful during development.