What are the different types of hooks available in Claude Code?

Claude Hooks supports three types: Command hooks for deterministic script execution, Prompt hooks for complex context-aware reasoning, and Agent hooks for multi-step verification tasks requiring tool access.

Can I use hooks to block dangerous terminal commands?

Yes, by utilizing the PreToolUse event with a matcher for the Bash tool, you can inspect command inputs and block execution by returning exit code 2 if a security policy is violated.

How do I debug a hook that isn't firing?

You can run Claude Code with the --debug flag to see execution logs, verify your matcher syntax, or manually pipe JSON input into your hook script to test its logic.

Where are hook configurations stored?

Hooks can be defined in .claude/settings.json for project-level sharing, ~/.claude/settings.json for personal global hooks, or within a plugin's hooks.json file for portable distributions.

Do multiple hooks run sequentially or in parallel?

All matching hooks for a specific event run in parallel rather than sequentially. This ensures performance but requires hooks to be designed for independent execution.

Claude Code Hooks & Workflow Automation

Name: Claude Code Hooks & Workflow Automation
Author: outfitter-dev

byoutfitter-dev

•

生産性とワークフロー

Automates development workflows and enforces security policies by intercepting Claude Code lifecycle events.

概要

The claude-hooks skill provides a comprehensive framework for extending Claude Code through event-driven triggers. It allows developers to create custom validation gates, automated formatting routines, and complex multi-step verifications by hooking into lifecycle events like tool execution, session starts, and subagent spawns. By supporting command-based scripts, LLM-driven prompt hooks, and tool-enabled subagent hooks, this skill enables a highly controlled and automated AI development environment tailored to specific project requirements and security standards.

主な機能

Configurable security gates to validate or block high-risk operations in real-time
Integration with system environment variables and project-specific configurations
Event-driven triggers for tool usage, session lifecycle, and subagent events
18 GitHub stars
Advanced regex-based matchers for precise tool and file pattern targeting
Support for Command, Prompt, and Subagent hook execution types

ユースケース

Automatically running linters, formatters, or type-checkers after Claude edits a file
Blocking dangerous shell commands or sensitive path access using PreToolUse validation
Initializing complex project environments and environment variables upon session start

概要

主な機能

Configurable security gates to validate or block high-risk operations in real-time
Integration with system environment variables and project-specific configurations
Event-driven triggers for tool usage, session lifecycle, and subagent events
18 GitHub stars
Advanced regex-based matchers for precise tool and file pattern targeting
Support for Command, Prompt, and Subagent hook execution types

ユースケース

Automatically running linters, formatters, or type-checkers after Claude edits a file
Blocking dangerous shell commands or sensitive path access using PreToolUse validation
Initializing complex project environments and environment variables upon session start