[BUG] Skill being accidentaly parsed as bash command

Preflight Checklist

• [x] I have searched existing issues and this hasn't been reported yet
• [x] This is a single bug report (please file separate reports for different bugs)
• [x] I am using the latest version of Claude Code

What's Wrong?

Claude seems to be parsing my skill as an attempted Bash command.
When attempting to load my example skill, I get an error.

What Should Happen?

The skill should be loaded properly with the examples.

Error Messages/Logs

/effective-effect
  ⎿  Error: Bash command permission check failed for pattern "!`
       return Console.log(shout ? message.toUpperCase() : message)
     })

     const cli = Command.run(greet, { name: "greet", version: "1.0.0" })

     cli(process.argv).pipe(Effect.provide(BunContext.layer), BunRuntime.runMain)
     `": Command contains newlines that could separate multiple commands

Steps to Reproduce

1. Add my skill from the following unpatched version: https://github.com/stromseng/skills/tree/bba727154e841928830f934cfb14e8c864e20a84
2. Run the skill in claude code using /effective-effect

Claude Model

Opus

Is this a regression?

I don't know

Last Working Version

_No response_

Claude Code Version

2.1.31

Platform

Anthropic API

Operating System

macOS

Terminal/Shell

Other

Additional Information

I fixed it by removing ! in the following commit:
https://github.com/stromseng/skills/commit/1a8a82b77a32d3372b765dce884d6b3826e49f9a