[BUG] Agent view: slash commands treated as prompts and forwarded to subagent

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 agent view is active (default since ~v2.1.139), typing a slash command such as /logout does not invoke the CLI command. Instead, the input is treated as a prompt and forwarded to the active subagent.

What Should Happen?

Slash commands should be intercepted and handled by the CLI regardless of whether agent view is active.

Error Messages/Logs

No error output. The command silently misbehaves rather than failing visibly.

Steps to Reproduce

1. Use Claude Code v2.1.139 or later (agent view is default)
2. At the prompt, type /logout (or any other slash command)
3. Observe that the command is sent as a prompt to the subagent rather than handled by the CLI

Claude Model

Sonnet (default)

Is this a regression?

Yes, this worked in a previous version

Last Working Version

2.1.138

Claude Code Version

2.1.152

Platform

Anthropic API

Operating System

macOS

Terminal/Shell

iTerm2

Additional Information

_No response_