[BUG] Claude Code Slash Commands No Longer Work

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?

When I run a /command with text that is a task then Claude Code will not run the command. It will just go rogue and start operating on the input to the command. This is because commands were changed to skills, so the agent has to run a tool to load the command rather than the CLI running them. As a result the agent gets confused every single time and wastes tokens running the command dynamically. Even if I explicitly tell it to run the command it won't do it. I have to tell it in following prompts after multiple attempts.

What Should Happen?

The command should run regardless of what comes after it.

Error Messages/Logs

Steps to Reproduce

Put in a task to a slash command like

/my-command Implement a countdown timer to New Year's as an HTML web page

The agent will not execute the command. It will start implementing the HTML web page.

Claude Model

Opus

Is this a regression?

Yes, this worked in a previous version

Last Working Version

_No response_

Claude Code Version

2.1.22

Platform

Anthropic API

Operating System

Windows

Terminal/Shell

WSL (Windows Subsystem for Linux)

Additional Information

<img width="3723" height="1089" alt="Image" src="https://github.com/user-attachments/assets/7e467b5b-5c53-46a3-a607-9f57a7d74c2c" />