[BUG] Claude code asking for permission to edit its own settings when bypass permissions is enabled

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?

I have an instance of claude code with bypass permissions on.
(It's running inside a dedicated VM to mitigate any security concerns)

Before this version of claude code, this meant I'd never be prompted to approve what claude was trying today.
Now on 2.1.78, I get prompted when claude code tries to edit its own settings (anything in ~/.claude)
I don't get prompted when it tries to run other commands. (as expected)

What Should Happen?

When bypass permissions mode is on, I expect to never be asked to approve an action the agent is making.

Error Messages/Logs

Steps to Reproduce

1. Configure claude code with the following settings.json:

{
  "env": {
    "CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS": "1"
  },
  "permissions": {
    "defaultMode": "bypassPermissions"
  },
  "skipDangerousModePermissionPrompt": true
}

1. Start claude
2. Ask claude to edit its own instructions

Claude Model

None

Is this a regression?

Yes, this worked in a previous version

Last Working Version

2.1.77

Claude Code Version

2.1.78 (Claude Code)

Platform

Anthropic API

Operating System

Ubuntu/Debian Linux

Terminal/Shell

Terminal.app (macOS)

Additional Information

_No response_