How do I prevent bash scripts from being corrupted in a skill?

You should use the pre-declaration pattern: capture arguments into named XML tags at the very top of your SKILL.md and reference those tags instead of using bare shell variables like $1 in your code blocks.

Can I use this skill as a template for my own projects?

Yes, this skill is specifically designed as a reference and test harness to help you verify substitution behavior before applying patterns to your production skills.

What is argument substitution in Claude Code?

Argument substitution is the process where Claude replaces variables like $0, $1, or $ARGUMENTS with the actual values provided by the user when invoking a skill via the terminal.

Why does my code disappear when I use $1 in a skill description?

If a skill takes arguments, Claude attempts to substitute $1 with the first argument. If no argument is provided, the variable is replaced with an empty string, effectively 'corrupting' your code example.

Argument Substitution Guide

Name: Argument Substitution Guide
Author: Jamie-BitFlight

byJamie-BitFlight

•

学习与文档

Demonstrates and tests argument substitution patterns to prevent code corruption in custom Claude Code skills.

This skill serves as a living test harness and architectural blueprint for developers building custom Claude Code skills. It provides a standardized method for capturing CLI arguments into named XML tags, ensuring that positional variables like $1 or $ARGUMENTS do not accidentally corrupt shell scripts or code blocks during execution. By utilizing this pre-declaration pattern, developers can safely include complex bash or awk examples in their skill documentation while maintaining clean, predictable access to user-provided parameters.

主要功能

01Live test harness for verifying $ARGUMENTS substitution behavior

0238 GitHub stars

03Visual demonstrations of common variable corruption pitfalls in bash and awk

04Implementation guide for safe code example referencing

05Standardized routing pattern for building argument-driven skill logic

06Positional argument capture ($0-$9) into structured XML tags

使用场景

01Testing how Claude Code handles complex CLI argument strings before deployment

02Debugging argument-passing logic for multi-parameter automation tasks

03Learning best practices for preventing shell variable corruption in custom skills

主要功能

01Live test harness for verifying $ARGUMENTS substitution behavior

0238 GitHub stars

03Visual demonstrations of common variable corruption pitfalls in bash and awk

04Implementation guide for safe code example referencing

05Standardized routing pattern for building argument-driven skill logic

06Positional argument capture ($0-$9) into structured XML tags

使用场景

01Testing how Claude Code handles complex CLI argument strings before deployment

02Debugging argument-passing logic for multi-parameter automation tasks

03Learning best practices for preventing shell variable corruption in custom skills