Slash commands for CLI vibecoding clients like Claude Code
Find a file
2025-08-07 18:06:51 +01:00
framework.md Initial commit 2025-08-04 00:18:37 +01:00
LICENSE Initial commit 2025-08-04 00:18:37 +01:00
README.md Initial commit 2025-08-04 00:18:37 +01:00

💻 Claude Code Commands

Slash commands for Claude Code and similar CLI coding assistants.

A collection of shortcuts and custom workflows accessible via / commands.

  1. What are Slash Commands?
  2. 🔧 Compatibility
  3. 📦 Setup
  4. 🚀 Usage
    1. Creating Custom Commands
    2. Best Practices for Command Design
  5. 📄 License

What are Slash Commands?

Slash commands are shortcuts that trigger specific behaviours in your CLI coding assistant. Type / followed by a command name to execute pre-configured instructions.

When you invoke a command, it:

  • Loads custom instructions from the corresponding .md file
  • Configures the assistant's behaviour for that specific task
  • Applies any specialised prompts or constraints
  • Executes the workflow you've defined

🔧 Compatibility

Designed for Claude Code and CLI coding assistants that support Markdown-based slash commands.

Note: As of the time of writing, the Gemini CLI does not have a similar slash command feature.

📦 Setup

  1. Fork this repository to your own Git account

  2. Clone to ~/.claude/commands:

    git clone ssh://git@git.tomfos.tr/tom/claude-code-commands.git ~/.claude/commands
    

    This gives you:

    • Your own command collection to customise
    • Version control for your modifications
    • Easy syncing across machines

🚀 Usage

Each .md file defines an available command. The filename determines the command name.

Example: write-docs.md/write-docs

Usage in Claude Code:

  1. Type / followed by the command name
  2. Press enter to load those instructions
  3. The assistant adopts the command's behaviour

Creating Custom Commands

Add a Markdown file with your instructions:

# my-command.md
You are an expert at [specific task]. When invoked:
1. [First action]
2. [Second action]
3. [Report results]

Best Practices for Command Design

  1. Single Purpose: Each command should do one thing well
  2. Clear Instructions: Be specific about what the command does
  3. Include Examples: Show typical usage scenarios
  4. Handle Edge Cases: Anticipate what might go wrong

📄 License

This project is licensed under the Apache License 2.0.