Claude Code Slash Commands: Reusable Prompts for Every Workflow

Slash commands are short, reusable prompts you invoke inside Claude Code by typing `/name`. They live as Markdown files in `.claude/commands/` and let you codify your team's playbook so every session speaks the same language.

Updated: April 24, 2026

Anatomy of a slash command

A `.md` file with YAML frontmatter (name, description, optional args) and a body that is the prompt itself. Arguments are interpolated as `$1`, `$2`, etc. That's it. The CLI picks it up on next start.

Commands we ship on every Alher Tech project

/new-feature (scaffolds backend + frontend + migration for a new entity), /add-stripe-plan (creates a plan in Stripe and wires the backend), /ship-pr (runs lint + test + opens PR), /fix-broken-test (triages a failing test and fixes or documents it), /security-review (runs our security checklist against the current branch).

Sharing commands across repos

Keep an agency-level commands library in its own git repo and symlink it into every project as `.claude/commands/`. Update once, roll out everywhere. That's how we keep 30+ active client SaaS projects consistent.

Frequently asked questions

Can slash commands accept arguments?

Yes. Reference them as `$1`, `$2` or `$ARGUMENTS` in the body. Typed args with descriptions also appear in autocomplete.