Bundled Skills

Bundled skills ship inside Claude Code, so you need no files to use them. They are prompts that guide Claude through a common job. For example, /simplify reviews changed code and fixes the problems it finds.

You cannot change a bundled skill. You can replace one with your own skill of the same name. Turn them off with the disableBundledSkills setting or CLAUDE_CODE_DISABLE_BUNDLED_SKILLS=1.

Example

claude -p "/debug the login test fails on CI"

/debug enables debug logging. It then investigates the issue that you describe.

Controls

• The setting disableBundledSkills and the env var CLAUDE_CODE_DISABLE_BUNDLED_SKILLS remove them. /doctor stays available.
• When the policy allow_bundled_skills is not granted, only names in a fixed list stay. The list in the program is: artifact-diagramming, batch, claude-code-docs, code-review, commit, dataviz, debug, explain-usage, fewer-permission-prompts, keybindings-help, loop, memory-types, plugin-authoring, pr, simplify, update-config, verify, workflow-authoring.

Skills

Skill	Status	What it does	Notes
/doctor (alias /checkup)	Documented: An official docs page names this item.	Checks your setup: install, unused extensions, CLAUDE.md, hooks, permissions. /doctor prompt-audit audits instructions.	Hidden by env var DISABLE_DOCTOR_COMMAND or skillOverrides "doctor": "off". Needs a workspace. Interactive.
/claude-api	Documented: An official docs page names this item.	A reference for the Claude API and Managed Agents in your language.	Subcommands: migrate, upgrade, api-version (new in 2.1.307: moves Claude API code to API version 2026-10-21, per the release notes), managed-agents-onboard, prompt-audit, cost-optimize, build-eval, hillclimb, preserved-thinking-migration, managed-agents-secure. Since 2.1.310, when two server flags (tengu_cosmic_sketch and tengu_typed_trinket) turn it on, it also searches the project for deprecated or retired model IDs and lists them in a section "Deprecated or Retired Models in This Project" (the program). The 2.1.311 release diff lists /claude-api as removed, but the skill is still in 2.1.311: the startup code still registers it and CLAUDE_CODE_DISABLE_CLAUDE_API_SKILL still turns it off (chunk-nxczhwwr.js:5483), and its definition is in chunk-cqcx451p.js:325. Only the way the program registers it changed, so the diff tool missed it. The built-in plugin cc-plugin-claude-platform also offers the same skill, only when no bundled skill has that name, and adds a docs MCP server (chunk-j1r347qy.js:160, chunk-j1r347qy.js:183).
/claude-api managed-agents-secure	Documented: An official docs page names this item.	Compares a Managed Agents agent's settings with Anthropic's published security guide. It changes nothing.	New in 2.1.304. It is unavailable when Claude Code's check on its ant commands does not run. Source: the release notes and the program.
/debug [issue description]	Documented: An official docs page names this item.	Enables debug logging and investigates.	disableModelInvocation. Allowed tools: Read, Grep, Glob.
/batch <instruction>	Documented: An official docs page names this item.	Plans a large change. Runs 5 to 30 worktree agents that open pull requests.	disableModelInvocation.
/run	Documented: An official docs page names this item.	Launches your app so you can see a change work.	Reads a recorded run-<name> skill if one exists.
/run-skill-generator	Documented: An official docs page names this item.	Records how to run the project as .claude/skills/run-<name>/.	disableModelInvocation.
/verify	Documented: An official docs page names this item.	Tries your change and reports if it works.	Records its recipe in .claude/skills/verify/SKILL.md. Since 2.1.312, /help and the / command menu show the short text "Try your change and report if it works" (chunk-ktc61wsq.js:5169). Claude gets the long description (chunk-ktc61wsq.js:5163). In 2.1.311 the menu showed the long description (chunk-nxczhwwr.js:5393). Source: the release notes and the program.
/code-review (alias /review), /loop, /simplify, /workflow-authoring	Documented: An official docs page names this item.	Named in the docs.	/workflow-authoring needs dynamic workflows.
/plugin-authoring	Documented: An official docs page names this item.	Helps you write and share a mod or plugin.	Since 2.1.304, sharing no longer stops at the secret scan on machines with managed settings or on Team and Enterprise plans. Source: the release notes.
/update-config	Documented: An official docs page names this item.	Edits settings.json: hooks, permissions, environment variables.	Registered as a bundled skill.
/fewer-permission-prompts	Documented: An official docs page names this item.	Scans transcripts and adds a read-only allowlist to .claude/settings.json.	Needs a workspace.
/security-review	Documented: An official docs page names this item.	Reviews pending branch changes for security risks.	A prompt command that Claude can also call through the Skill tool. Git repositories.
/schedule (alias /routines)	Documented: An official docs page names this item.	Creates and manages scheduled remote agents.	A stub with a name, alias, and description. Needs claude.ai.
/keybindings-help	Undocumented: Only the program shows this item. It can change without notice.	Helps you edit ~/.claude/keybindings.json.	userInvocable:false, so only Claude runs it. Allowed tool: Read.
/claude-in-chrome	Gated: Works only behind a flag or a server gate. Needs the Chrome extension, or a check passing.	Browser automation through the Chrome extension.	No allowed tools.
/design-sync [<project hint>]	Gated: Works only behind a flag or a server gate. Policy allow_design_sync.	Pushes design-system components to claude.ai/design.	disableModelInvocation.
/commit	Undocumented: Only the program shows this item. It can change without notice. Gated: Works only behind a flag or a server gate. An isEnabled check decides.	Creates a git commit. It gathers git context first.	A prompt command.
/pr	Undocumented: Only the program shows this item. It can change without notice. Gated: Works only behind a flag or a server gate. An isEnabled check decides.	Creates a GitHub pull request. It gathers branch context first.	A prompt command.
/commit-push-pr	Undocumented: Only the program shows this item. It can change without notice.	Makes a commit. Pushes it. Opens a pull request.	A prompt command.
/dataviz	Undocumented: Only the program shows this item. It can change without notice.	Guides Claude when it makes a chart, graph, plot, dashboard, or other data visualization.	A prompt command.
/explain-usage	Undocumented: Only the program shows this item. It can change without notice. Gated: Works only behind a flag or a server gate. An isEnabled check decides.	Explains where the tokens of this session went, with one simple chart.	A prompt command.
/memory-types	Undocumented: Only the program shows this item. It can change without notice. Gated: Works only behind a flag or a server gate. An isEnabled check decides.	A full reference for the memory types: what each captures, when to save it, and how to write the body.	A prompt command.
/artifact-capabilities	Undocumented: Only the program shows this item. It can change without notice. Gated: Works only behind a flag or a server gate. An isEnabled check decides.	A guide for Artifacts. The program gives no description.	A prompt command.
/artifact-design	Undocumented: Only the program shows this item. It can change without notice. Gated: Works only behind a flag or a server gate. An isEnabled check decides.	Design guidance and fundamentals for Artifacts.	A prompt command.
/artifact-diagramming	Undocumented: Only the program shows this item. It can change without notice. Gated: Works only behind a flag or a server gate. An isEnabled check decides.	Diagramming guidance for Artifacts: when a picture helps, and how to draw it with inline SVG in both themes.	A prompt command.
/design	Undocumented: Only the program shows this item. It can change without notice. Gated: Works only behind a flag or a server gate. An isEnabled check decides.	A design command. The program gives no description.	A prompt command.
/claude-code-docs	Undocumented: Only the program shows this item. It can change without notice. Gated: Works only behind a flag or a server gate. An isEnabled check decides, and it reads the server flag tengu_birch_kettle.	A command for questions about Claude Code. The program gives no description.	A prompt command.
/cowork-plugin	Undocumented: Only the program shows this item. It can change without notice. Gated: Works only behind a flag or a server gate. An isEnabled check decides.	Creates a new Cowork plugin from scratch, or customizes an installed plugin for an organization.	A prompt command.
/setup-claude (alias /setup-cowork)	Undocumented: Only the program shows this item. It can change without notice. Gated: Works only behind a flag or a server gate. An isEnabled check decides.	A guided setup. You pick a role, install a matching plugin, try a skill, and connect tools.	A prompt command.

Related Pages

• Skills and Commands: where skills live and how to turn them off.
• Slash Commands: the built-in commands.
• Skill Frontmatter: the keys named above, such as disable-model-invocation.

Source: Commands, Extend Claude with Skills, and the program (the list of bundled skill names and the allow_bundled_skills key).