Short answer
Claude Code skills are reusable instruction packages for Claude workflows. They help Claude do a specific job, such as review code, write tests, or update docs.
Words used in this guide
- AIPM: a tool that installs and updates reusable AI instructions.
- Package: a named group of files that people can install and update together.
- Workflow: a set of steps used to finish a task.
- Repository or repo: a project folder whose changes are saved and tracked.
- MCP: a standard way for an AI tool to connect to other tools and information.
What this means
Skills are for repeatable work
Use a skill when the same task happens often. Good examples are code review, issue triage, release notes, test writing, and migration steps. Not sure a skill is the right tool? Compare skills vs MCP vs subagents vs hooks.
A skill should be focused
A small skill is easier to trust. It should say what input it needs, what output to produce, and what limits to follow.
AIPM helps distribute skills
AIPM gives each skill a name and version number. This helps a team install the same skill in many projects. For every folder Claude Code loads skills from, see where are Claude skills stored?.
Simple steps
- Pick one repeated Claude Code task.
- Write the instructions and expected output.
- Add examples from real project work.
- Test the skill in one repo.
- Publish and install it with AIPM when it is ready.
Where to go next
If you want to try this in a real project, start with how to use AIPM and the command reference. If you want to share your own workflow, read the publishing guide.
FAQ
Is a Claude Code skill just a prompt?
It can include prompt-like instructions, but it is packaged as a reusable workflow instead of a one-time chat message.
Should every workflow become a skill?
No. Start with tasks that repeat and have a clear output format.