Can I customize which commands are blocked?

Yes, you can modify the patterns.yaml file to add specific bash patterns, CLI commands, or file paths that should be restricted or require manual confirmation.

Does it support Windows and Linux?

Yes, Damage Control includes workflows to handle Unix-based shells as well as Windows PowerShell and CMD patterns.

How do I install these security hooks?

You can initiate the setup by telling Claude to 'install damage control' or 'setup security hooks,' which will trigger the automated installation workflow.

What are the different path protection levels?

The system offers 'zeroAccess' (no access at all), 'readOnly' (modifications blocked), and 'noDelete' (all operations allowed except deletion).

What is the Damage Control skill for Claude Code?

Damage Control is a security skill that uses PreToolUse hooks to intercept and block potentially harmful commands or file modifications before they are executed by the AI.

Damage Control Security Hooks

Name: Damage Control Security Hooks
Author: bjornallvin

bybjornallvin

보안 및 테스팅

Protects your local environment by blocking dangerous bash commands and enforcing granular file access restrictions during AI interactions.

소개

Damage Control is a defense-in-depth security system designed specifically for Claude Code. It utilizes PreToolUse hooks to intercept potentially destructive actions before they are executed, providing a critical safety net for developers. By defining custom command patterns and path-specific access levels—such as zeroAccess for credentials and readOnly for system configs—users can safely delegate tasks to Claude without risking accidental data loss or system instability. It supports both Python and TypeScript runtimes and includes cross-platform compatibility for Unix and Windows environments.

주요 기능

Granular path protection levels including ZeroAccess, ReadOnly, and NoDelete
Dual runtime support for Python (UV) and TypeScript (Bun)
0 GitHub stars
Command pattern blocking for dangerous operations like recursive deletes and hard resets
Interactive 'Ask' triggers that require manual confirmation for risky-but-valid commands
Cross-platform support with specific protection patterns for Bash, PowerShell, and CMD

사용 사례

Preventing AI agents from accidentally deleting critical system configurations or credentials
Standardizing security policies across team projects using shared hook configurations
Restricting Claude's access to sensitive environment variables and .env files

소개

주요 기능

Granular path protection levels including ZeroAccess, ReadOnly, and NoDelete
Dual runtime support for Python (UV) and TypeScript (Bun)
0 GitHub stars
Command pattern blocking for dangerous operations like recursive deletes and hard resets
Interactive 'Ask' triggers that require manual confirmation for risky-but-valid commands
Cross-platform support with specific protection patterns for Bash, PowerShell, and CMD

사용 사례

Preventing AI agents from accidentally deleting critical system configurations or credentials
Standardizing security policies across team projects using shared hook configurations
Restricting Claude's access to sensitive environment variables and .env files