Should I use prompt-based or command hooks?

Use prompt-based hooks for complex reasoning and context-aware validation. Use command hooks for fast, deterministic checks like file system operations or external tool integrations.

Where do I store plugin hook configurations?

For individual plugins, hooks should be defined in a 'hooks/hooks.json' file using a specific wrapper format that includes a 'hooks' object containing the event definitions.

How do I secure my plugin hooks?

Security best practices include implementing strict input validation using jq, checking for path traversal in file operations, and always quoting variables in bash scripts to prevent injection.

Can hooks modify tool behavior?

Yes, PreToolUse hooks can specifically approve, deny, or even modify the inputs of a tool call before the execution takes place.

What are Claude Code hooks?

Hooks are event-driven automation scripts or prompts that execute in response to specific Claude Code events, such as tool calls, session starts, or agent completion.

Hook Development for Claude Code

Name: Hook Development for Claude Code
Author: SwervoJake

bySwervoJake

0•

开发者工具

Simplifies the creation and implementation of event-driven hooks for Claude Code plugins to automate workflows and enforce safety policies.

This skill provides comprehensive guidance for developers building Claude Code plugins, focusing on the powerful Hooks API. It enables the implementation of both prompt-based and command-based hooks, allowing for context-aware validation, security enforcement, and automated environment setup. Whether you need to block dangerous commands, validate tool inputs before execution, or load specific project context at the start of a session, this skill offers the patterns and configuration formats necessary to build robust, event-driven integrations within the Claude Code ecosystem.

主要功能

01Comprehensive event coverage including PreToolUse, PostToolUse, Stop, and SessionStart

02Integrated security patterns for path traversal prevention and input validation

03Standardized configuration formats for both plugin hooks.json and user settings

04Support for both LLM-driven prompt hooks and deterministic bash command hooks

05Environment variable persistence across sessions via $CLAUDE_ENV_FILE

060 GitHub stars

使用场景

01Blocking potentially dangerous bash commands or file modifications through automated validation

02Automatically loading project-specific environment variables and context when a session starts

03Enforcing code quality and completion standards by validating agent outputs before the session stops

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add swervojake/claaaude-studio hook-development

For use in Claude.ai and ChatGPT

主要功能

01Comprehensive event coverage including PreToolUse, PostToolUse, Stop, and SessionStart

02Integrated security patterns for path traversal prevention and input validation

03Standardized configuration formats for both plugin hooks.json and user settings

04Support for both LLM-driven prompt hooks and deterministic bash command hooks

05Environment variable persistence across sessions via $CLAUDE_ENV_FILE

060 GitHub stars

使用场景

01Blocking potentially dangerous bash commands or file modifications through automated validation

02Automatically loading project-specific environment variables and context when a session starts

03Enforcing code quality and completion standards by validating agent outputs before the session stops