Claude/Code/Skill

• Claude Code Skill

  • A Claude/Code skill is a SKILL.md file inside a named directory that extends what Claude can do, either through explicit invocation (/skill-name) or automatic discovery when relevant.

  • Command merger (2026)

    • Claude/Code merged custom commands into skills. A file at .claude/commands/deploy.md and .claude/skills/deploy/SKILL.md both create /deploy and work identically.
    • .claude/commands/ files continue to work, but skills are the recommended format — they support supporting files, frontmatter invocation control, and automatic discovery by Claude.
    • This parallels Codex/CLI deprecating custom prompts in favor of Agent/Skills in codex-cli 0.117.0. See Codex/CLI/Q/Why doesn't CodexCLI support slash commands?

  • Open standard

    • Claude Code skills follow the Agent/Skills open standard (agentskills.io), which works across multiple AI tools including Codex/CLI.
    • Claude Code extends the standard with invocation control, subagent execution, and dynamic context injection.

  • Structure

    • skill-name/
      ├── SKILL.md        # required entrypoint
      ├── reference.md    # optional supporting docs
      └── scripts/
          └── helper.sh   # optional scripts

    • Each skill directory goes under a scope path; the directory name becomes the /command.

  • Scope levels

    • Level	Path	Applies to
      Personal	~/.claude/skills/<name>/SKILL.md	All projects
      Project	.claude/skills/<name>/SKILL.md	This project only
      Plugin	<plugin>/skills/<name>/SKILL.md	Where plugin is enabled

  • Key frontmatter fields

    • Field	Purpose
      description	What the skill does; Claude uses this for auto-discovery
      when_to_use	Additional trigger phrases appended to description
      disable-model-invocation	true = only you can invoke (deploy, commit, etc.)
      user-invocable	false = only Claude can invoke (background knowledge)
      allowed-tools	Tools Claude can use without per-use approval when skill is active
      context	fork = run in an isolated subagent
      agent	Which subagent type to use when context: fork is set
      argument-hint	Hint shown during autocomplete, e.g. [issue-number]
      paths	Glob patterns that limit when skill auto-activates
      model	Model override for the duration of the skill

  • Dynamic context injection

    • Use !`command` in skill content to run shell commands before Claude sees the prompt — output replaces the placeholder inline.
    • Multi-line variant: fenced block opened with `! .

  • Invocation control summary

    • Frontmatter	You can invoke	Claude can invoke
      (default)	Yes	Yes
      disable-model-invocation: true	Yes	No
      user-invocable: false	No	Yes

  • Docs

    • Extend Claude with skills – Claude Code