Hooks are user-defined shell commands that execute at specific points in Claude Code's lifecycle. They provide control over Claude's behavior, ensuring certain actions always happen rather than relying on the AI to choose to run them.

Files in this Directory

format-after-edit.sh - A PostToolUse hook that automatically formats code after file edits
example-hook-config.json - Example configuration showing how to set up various hooks

How to Use These Hooks

Option 1: Copy to Your Settings File

Copy the hooks configuration from example-hook-config.json to your Claude Code settings:

Project-specific (.claude/settings.json):

# Create settings file if it doesn't exist
touch .claude/settings.json

# Add hooks configuration from example-hook-config.json

User-wide (~/.claude/settings.json):

# Apply hooks to all Claude Code sessions
cp example-hook-config.json ~/.claude/settings.json

Option 2: Use Individual Hooks

Copy the hook script to your project:

cp format-after-edit.sh /your/project/.claude/hooks/
chmod +x /your/project/.claude/hooks/format-after-edit.sh

Add to your settings.json:

{
  "hooks": {
    "PostToolUse": [
      {
        "matcher": "Edit|Write|MultiEdit",
        "hooks": [
          {
            "type": "command",
            "command": ".claude/hooks/format-after-edit.sh"
          }
        ]
      }
    ]
  }
}

Available Hook Events

PreToolUse: Before tool execution (can block tools)
PostToolUse: After successful tool completion
UserPromptSubmit: When user submits a prompt
SubagentStop: When a subagent completes
Stop: When main agent finishes responding
Notification: During system notifications
PreCompact: Before context compaction
SessionStart: At session initialization

Creating Your Own Hooks

Write a shell script that:
- Reads JSON input from stdin
- Processes the input
- Returns JSON output (empty {} for success)
- Can return {"action": "block", "message": "reason"} to block operations
Make it executable:

chmod +x your-hook.sh

Add to settings.json with appropriate matcher and event

Security Considerations

Hooks execute arbitrary shell commands
Always validate and sanitize inputs
Use full paths to avoid PATH manipulation
Be careful with file operations
Test hooks thoroughly before deployment

Debugging Hooks

Run Claude Code with debug flag to see hook execution:

claude --debug

This will show:

Which hooks are triggered
Input/output for each hook
Any errors or issues

Integration with Subagents

The example configuration includes a hook that integrates with the validation-gates subagent, demonstrating how hooks and subagents can work together for a more robust development workflow.