Claude Code
Claude Code skills: instructions that load when needed
A Claude Code skill is a folder with a SKILL.md file that holds a name, a description and instructions. The description is always in context, so Claude knows which skills it has. The rest loads only when you run it with /name or when your request matches the description.
Updated
What is a skill in Claude Code?
A folder with a SKILL.md: YAML frontmatter with name and description, then the instructions in Markdown. Reference files and scripts can sit next to it. Claude Code follows Agent Skills, an open format created by Anthropic, and extends it.
How does Claude know when to load a skill?
- The name and description of every skill are always in context.
- The body of SKILL.md enters context only when the skill is used.
- Reference files are read when the instructions point to them. Scripts run, and only their output reaches the context.
- With disable-model-invocation, Claude won't start the skill on its own. You start it with /name.
When is a skill worth writing?
When you keep pasting the same instructions into chat, or when a section of CLAUDE.md has turned into a procedure. CLAUDE.md loads in every session and a skill only when it is used, so long procedures take no context the rest of the time. One example: a release skill with the steps written out and a script that checks the version.
Skill or subagent?
A skill brings instructions into the current conversation. A subagent is a separate assistant with its own context that returns a summary. You can combine them: with context: fork, the skill runs in a subagent and does not see the conversation history.
Is a downloaded skill safe?
- Anthropic recommends skills only from trusted sources: ones you wrote or ones from Anthropic.
- A malicious skill can call tools or run code in ways its description doesn't mention, including to send data out.
- The allowed-tools field grants permissions. Check it on skills inside a repository before you run Claude Code there.
- Read everything in the folder, scripts included.
Frequently asked questions
Skill or slash command?
Custom commands have been merged into skills: a command and a skill with the same name both create /name. Skills add supporting files, control over invocation and automatic loading.
What goes in a skill's description?
What the skill does and when to use it. Claude matches your request against the description to decide whether to start the skill.
Where do skills live?
In the project, so the team gets them through the repository, in your personal folder for all your projects, or in a plugin.
Do skills work outside Claude Code?
Yes. Skills also work through the Claude API and in claude.ai, but each surface loads them separately.
Sources
Related articles
Claude Code subagents: what they are and when to use them
A Claude Code subagent is a separate assistant with its own context window, its own system prompt and only the tools you allow it. Claude hands it a task, the subagent works on its own and sends back just a summary, so long logs and search results stay out of your main conversation.ReadWhat is MCP (Model Context Protocol) and what an MCP server does
MCP (Model Context Protocol) is an open-source standard that connects AI applications to external systems such as files, databases, tools and services. An MCP server exposes tools, resources and prompts, and the application, for example Claude Code, opens one client for each server it connects to.Read
Case studies
Related services
Want skills for your team's procedures?
Tell us which steps your team repeats most often. We will work out together what can be written as a skill and what should stay with a person.
Write to us