Complete list of built-in skills and slash commands available in Claude Code — with descriptions and when to use each.
What are Claude Code slash commands? Slash commands (also called "skills") are specialized instruction sets that Claude Code loads on demand. Type /command-name at the start of any message to trigger a focused workflow — code review, security audit, automation, and more. As of May 2026, Claude Code ships with 22 built-in commands.
Pull Request Review — Deep multi-file analysis. Surfaces logic bugs, security issues, style concerns, and missed edge cases. Use before merging any non-trivial PR.
Multi-Agent Cloud Review — Parallel review by multiple specialized agents. Use for /ultrareview to review current branch, or /ultrareview 123 to review GitHub PR #123.
Security Audit — Focused OWASP Top 10 check: injection flaws, auth bypasses, unsafe deps. Use before any deploy touching auth, payments, or user data.
Code Simplification — Reviews recently changed files for unnecessary complexity, then rewrites — same logic, cleaner code.
Targeted Refactor — Extracts helpers, removes duplication, improves naming. Behavior-preserving only.
Recurring Tasks — Runs a prompt or slash command on a repeating interval. Self-paced — Claude decides when to iterate next. Example: /loop 5m /review.
Cron Scheduling — Create scheduled remote agents that run on a cron schedule. One-off or recurring. Example: /schedule run this at 3pm.
Initialize CLAUDE.md — Scans your codebase and writes a tailored CLAUDE.md with project conventions, tech stack, and architecture overview.
Configure Settings — Updates settings.json for permissions, hooks, env vars, and allowed tools. Use for "allow X", "add permission", or "set DEBUG=true".
Keyboard Shortcuts — Customize Claude Code keybindings in ~/.claude/keybindings.json. Rebind keys, add chord shortcuts, change submit key.
Reduce Prompts — Scans your transcripts and adds a prioritized allowlist to settings to stop repeated permission popups for common safe operations.
Status Line — Configure the terminal status line showing model name, token usage, and cost indicators.
Debug Assistant — Systematic root-cause analysis from stack traces. Forms hypotheses and guides targeted fixes rather than guessing.
Write & Run Tests — Generates tests for happy path and edge cases, infers the test framework from the project, and runs the suite.
Implementation Planner — Maps affected files, proposes a step-by-step plan, and surfaces architectural trade-offs before writing any code.
Explain Code — Plain-English walkthrough of any code block, file, or system: intent, data flow, side effects, and gotchas.
Deploy Helper — Guides deployment steps for your stack: env vars, build scripts, infra config. Stack-aware.
Generate Changelog — Reads git log since the last tag and produces a structured CHANGELOG entry grouped by feature and fix.
Dependency Audit — Checks package.json / requirements.txt / Cargo.toml for outdated, vulnerable, or redundant dependencies.
Generate Docs — Writes JSDoc, docstrings, or README sections for a file. Infers types and edge cases from the implementation.
Port Code — Translates code between languages (Python↔TypeScript, Go↔Rust) while preserving logic and updating tests.
Claude API Assistant — Build, debug, and optimize Anthropic SDK apps with prompt caching, tool use, streaming, and model migrations.
The interactive browser lets you search by keyword, filter by category, and see detailed examples for each command.
Open Claude Skills Browser →/review or /security-review. Claude Code detects the skill name, loads its specialized instructions, and executes the workflow. Some commands accept arguments: /ultrareview 42 reviews GitHub PR #42.How to use /review → | Automating tasks with /loop and /schedule →