claude-code-hooks

name	claude-code-hooks
description	Implement Claude Code hooks for deterministic control over agent behavior. Use when creating custom hooks for notifications, auto-formatting, logging, feedback, permissions, or lifecycle events.

Claude Code Hooks

Hooks are shell commands that execute at specific points in Claude Code's lifecycle, providing deterministic control over behavior.

Hook Events

Event	When it runs	Can block?
PreToolUse	Before tool execution	Yes
PermissionRequest	When permission dialog shown	Yes
PostToolUse	After tool completes	Feedback only
UserPromptSubmit	When user submits prompt	Yes
Notification	When notification sent	No
Stop	When agent finishes	Can continue
SubagentStop	When subagent finishes	Can continue
PreCompact	Before compact operation	No
SessionStart	Session starts/resumes	No
SessionEnd	Session ends	No

Configuration

Hooks are defined in settings files:

• ~/.claude/settings.json - User settings (all projects)
• .claude/settings.json - Project settings
• .claude/settings.local.json - Local settings (not committed)

Basic Structure

{
  "hooks": {
    "EventName": [
      {
        "matcher": "ToolPattern",
        "hooks": [
          {
            "type": "command",
            "command": "your-command-here",
            "timeout": 60
          }
        ]
      }
    ]
  }
}

Matcher patterns:

• Exact match: Write, Bash, Read
• Regex: Edit|Write, Notebook.*, mcp__memory__.*
• Match all: * or ""

Hook Input

Hooks receive JSON via stdin:

{
  "session_id": "abc123",
  "transcript_path": "/path/to/transcript.jsonl",
  "cwd": "/current/directory",
  "permission_mode": "default",
  "hook_event_name": "PreToolUse",
  "tool_name": "Write",
  "tool_input": { "file_path": "/path/to/file", "content": "..." },
  "tool_use_id": "toolu_01ABC..."
}

Hook Output

Exit Codes

• 0: Success (stdout shown in verbose mode)
• 2: Blocking error (stderr fed back to Claude)
• Other: Non-blocking error (stderr shown in verbose mode)

JSON Output (exit code 0)

{
  "decision": "block",
  "reason": "Explanation for Claude",
  "continue": true,
  "stopReason": "Message when continue=false",
  "systemMessage": "Warning for user"
}

Quick Examples

Log Bash Commands

{
  "hooks": {
    "PreToolUse": [
      {
        "matcher": "Bash",
        "hooks": [{
          "type": "command",
          "command": "jq -r '.tool_input.command' >> ~/.claude/bash.log"
        }]
      }
    ]
  }
}

Auto-Format on Save

{
  "hooks": {
    "PostToolUse": [
      {
        "matcher": "Edit|Write",
        "hooks": [{
          "type": "command",
          "command": "jq -r '.tool_input.file_path' | xargs -I{} sh -c 'echo \"{}\" | grep -q \"\\.ts$\" && npx prettier --write \"{}\"'"
        }]
      }
    ]
  }
}

Block Sensitive Files

{
  "hooks": {
    "PreToolUse": [
      {
        "matcher": "Edit|Write",
        "hooks": [{
          "type": "command",
          "command": "jq -e '.tool_input.file_path | test(\"\\\\.env|secrets|credentials\")' > /dev/null && echo 'Cannot modify sensitive files' >&2 && exit 2 || exit 0"
        }]
      }
    ]
  }
}

Environment Variables

• CLAUDE_PROJECT_DIR - Absolute path to project root
• CLAUDE_PLUGIN_ROOT - Plugin directory (for plugin hooks)
• CLAUDE_ENV_FILE - File to persist env vars (SessionStart only)
• CLAUDE_CODE_REMOTE - "true" if running in remote/web environment

References

• Decision Control: See references/decision-control.md for PreToolUse, PermissionRequest, Stop hook decisions
• Event Details: See references/events.md for complete input/output schemas per event
• Examples: See references/examples.md for real-world hook implementations