# Agent Skills

Agent Skills are modular instruction sets that guide AI coding agents toward the right information at the right time. They automate [context engineering](/umbraco-in-ai/concepts/context-engineering.md) through active context management. Skills are not activated unless the task requires them. They use progressive discovery so the agent starts with a summary and pulls in deeper detail only when needed.

For Umbraco, skills allow us to:

* **Guide agents to the right information** — pointing them to current APIs and platform-specific patterns rather than letting them guess
* **Show best practice** — providing validated code examples and workflow guidance for each extension type
* **Deliver the right Umbraco information at the right time** — the agent only receives context relevant to the current task

## What's Inside a Skill?

Each skill is a directory containing a `SKILL.md` file along with supporting reference files and scripts:

* **`SKILL.md`** — the main instruction file with workflow guidance and code patterns
* **Reference files** — detailed documentation, style guides, or API references that the skill loads on demand
* **Scripts** — helper scripts the AI can run to validate work or automate checks

When you ask your AI assistant to "create a dashboard", it loads the `umbraco-dashboard` skill and gets precise, current instructions instead of guessing.

## What Skills Are Not

* **Not a replacement for documentation.** Skills point to the documentation and teach the AI how to apply it. You should still read the [Umbraco documentation](https://docs.umbraco.com/) yourself.
* **Not scaffolding tools.** They don't generate boilerplate from templates. They teach the AI to write code that fits your specific requirements.
* **Not perfect.** Skills improve results but cannot guarantee correctness. Always review generated code.

## How Skills Differ Across Editors

Skills use the open [`SKILL.md` format](https://agentskills.io/home), which is supported by multiple AI coding tools. The experience varies slightly:

| Editor             | How Skills Load                          | Installation                       |
| ------------------ | ---------------------------------------- | ---------------------------------- |
| **Claude Code**    | Native plugin system, on-demand          | `/plugin install` from marketplace |
| **Cursor**         | `.cursor/skills/` directory, on-demand   | `npx skills add` via Skills CLI    |
| **GitHub Copilot** | `.github/skills/` directory, on-demand   | `npx skills add` via Skills CLI    |
| **Windsurf**       | `.windsurf/skills/` directory, on-demand | `npx skills add` via Skills CLI    |

The skill content is identical across editors. What differs is how each editor discovers and loads the skill files. Claude Code uses a plugin/marketplace system; other editors use the [Vercel Skills CLI](https://github.com/vercel-labs/skills) to install skills into editor-specific directories.

Some advanced features (like spawning reviewer agents or entering plan mode) work best in Claude Code. The core value — correct code patterns from current documentation — works everywhere.

## Skills and MCP

Skills and the [Model Context Protocol (MCP)](/umbraco-in-ai/mcp/base-mcp.md) are complementary. Skills provide **knowledge** — they teach the agent how to write correct code. MCP provides **capability** — it gives the agent tools to interact with a running Umbraco instance. An agent might use a skill to learn how to build a dashboard, and use MCP tools to create the content types it needs.


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://docs.umbraco.com/umbraco-in-ai/concepts/agent-skills.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.