Version 1.0.72

What: New /statusline command for customizing Claude Code's status line UI Why: Allows users to personalize their terminal experience and display custom information in the status line How to use:

# Configure status line from your shell PS1
/statusline

# Configure with custom command
/statusline "Configure my custom status display"

Details:

• Creates a specialized "statusline-setup" agent to handle configuration
• Can read and convert existing shell PS1 configurations
• Saves custom scripts to ~/.claude/statusline-command.sh if needed
• Updates settings in ~/.claude/settings.json

What: Improved visual indicators for different Claude Code operational modes Why: Makes it clearer which mode Claude Code is operating in, improving user awareness and control How to use: The mode display is automatic and shows in the interface with distinct icons and colors:

• Default mode: "Default" - standard operation
• Plan mode: "Plan" with ⏸ icon - planning before execution
• Accept Edits mode: "Accept" with ⏵⏵ icon - auto-accepting file edits
• Bypass Permissions mode: "Bypass" with warning styling - security restrictions bypassed

Details:

• Color-coded indicators (text, planMode, autoAccept, error)
• Clear visual distinction between modes
• Better integration with status line

What: Enhanced validation for file system commands with clearer security messages Why: Provides better security boundaries and more helpful error messages when paths are restricted How to use: Path validation is automatic for commands like cd, ls, and find. When blocked, you'll see specific messages like:

cd in '/restricted/path' was blocked. For security, Claude Code may only 
change directories to the allowed working directories for this session: [list]

Details:

• Better handling of glob patterns (*, ?, [, ], {, })
• Proper tilde expansion (~ and ~/ paths)
• Improved handling of relative paths with ..
• More specific error messages for each command type

What changed: Better error reporting for invalid agent configuration files Impact: Clearer error messages when agent files are missing required fields or have invalid configurations, making it easier to debug custom agents

What changed: More accurate calculation of token usage including cache tokens Impact: Better cost tracking with separate accounting for cached vs uncached tokens, providing more accurate usage reports

What changed: New component for displaying running background shells in the UI Impact: Better visibility of background processes with indicators showing shell count and status

• Issue: Certain path patterns with glob characters or parent directory references were not properly validated
• Resolution: Improved regex patterns and path resolution logic to handle all edge cases correctly

• Issue: Unused imports from crypto and stream modules were left in the codebase
• Resolution: Removed unused imports and reorganized module dependencies for cleaner code

• Issue: Missing or invalid frontmatter fields in agent files produced unclear error messages
• Resolution: Added specific validation and error reporting for required fields (name, description, model)