create-slash-commands_skill

This skill helps you create Claude Code slash commands that trigger reusable prompts with structured XML and YAML frontmatter for consistency.
  • Shell

1.2k

GitHub Stars

1

Bundled Files

2 months ago

Catalog Refreshed

4 months ago

First Indexed

Readme & install

Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.

Installation

Preview and clipboard use veilstrat where the catalogue uses aiagentskills.

npx veilstrat add skill glittercowboy/taches-cc-resources --skill create-slash-commands

  • SKILL.md15.9 KB

Overview

This skill teaches how to author effective Claude Code slash commands that expand into reusable prompts with /command-name syntax. It covers YAML frontmatter, required XML tags, dynamic context loading, argument handling, and when to restrict tools. Followable rules and examples make it easy to create consistent team or personal commands.

How this skill works

You create markdown files in .claude/commands/ (project) or ~/.claude/commands/ (personal) with YAML frontmatter and an XML-tagged body. The skill enforces required tags (<objective>, <process>, <success_criteria>) and shows when to add <context>, <verification>, <testing>, or <output>. It explains argument patterns ($ARGUMENTS, $1/$2) and how to run shell commands into the prompt using !... for dynamic state.

When to use it

  • Standardize repeated prompts across a team or project
  • Create single-step checks or analysis utilities (no artifacts)
  • Create multi-step workflows that produce files, commits, or PR reviews
  • Build parameterized commands that accept issue numbers, file paths, or positional args
  • Add context-aware commands that load git status, diffs, or file contents

Best practices

  • Always use XML structure after frontmatter: <objective>, <process>, <success_criteria> are mandatory
  • Use clear, human-friendly description in frontmatter (shown in /help)
  • Include argument-hint and reference $ARGUMENTS or positional $1/$2 when user input is required
  • Use dynamic context (!command) and file references (@ path) for stateful tasks
  • Restrict allowed-tools in frontmatter for any command that runs shell actions or modifies files
  • When commands produce files, include <output> and <verification> with concrete checks

Example use cases

  • /optimize src/utils/helpers.js — analyze a file for performance and suggest three fixes
  • /fix-issue 123 — apply coding standards and add tests for a specific issue number
  • /create-commit — load git status, craft a commit message, and perform a commit (with allowed-tools restricted)
  • /check-todos — scan known TO-DOS.md and report remaining items and owners
  • /review-pr 456 high alice — positional-argument example to review PR #456 with priority and assignee

FAQ

Put files named command-name.md in .claude/commands/ for project commands or ~/.claude/commands/ for personal commands. Invoke them with /command-name in a conversation.

How do I decide if a command needs arguments?

If the task operates on user-specified data (issue numbers, file paths, PRs) include argument-hint and use $ARGUMENTS or positional variables. If it uses implicit context (current conversation, known files), omit argument-hint.

Built by
VeilStrat
AI signals for GTM teams
© 2026 VeilStrat. All rights reserved.All systems operational