Does this replace the hook-scope-guide?

No, it works alongside it. While hook-scope-guide helps you decide where to place hooks, hooks-eval focuses on auditing, testing, and implementing them correctly.

What security issues does the framework detect?

It scans for dangerous code patterns, potential injection vulnerabilities, and ensures proper data validation and sanitization within your hook logic.

What types of hooks can this skill evaluate?

It supports both JSON-based hooks and programmatic hooks using the Claude Agent SDK (Python), covering plugin, project, and global scopes.

Can I use this to fix performance issues in my AI agents?

Yes, it provides performance benchmarking and optimization guidance to identify hooks that are causing high latency or slowing down your execution cycle.

Claude Hook Evaluator

Name: Claude Hook Evaluator
Author: athola

byathola

•

Security & Testing

Evaluates and audits Claude Code hooks for security vulnerabilities, performance bottlenecks, and SDK compliance.

About

The hooks-eval skill provides a robust framework for developers to audit and optimize Claude Code and Agent SDK hooks. It streamlines the process of scanning for security vulnerabilities like injection patterns, benchmarking execution performance, and ensuring architectural compliance across plugin, project, and global scopes. Whether you are implementing programmatic Python SDK hooks or JSON-based rules, this tool offers a standardized scoring system and quality gates to ensure your AI agents operate safely, reliably, and efficiently.

Key Features

10 GitHub stars
Comprehensive SDK reference for Python-based hook implementations
Automated security scanning for vulnerability and injection prevention
Compliance validation against hook architecture best practices
Standardized quality scoring system with detailed reporting
Performance benchmarking for execution time and resource usage

Use Cases

Auditing third-party Claude Code plugins for security risks before deployment
Optimizing slow-performing hooks to reduce latency in AI agent workflows
Ensuring new hook implementations follow official Agent SDK callback patterns

About

Key Features

10 GitHub stars
Comprehensive SDK reference for Python-based hook implementations
Automated security scanning for vulnerability and injection prevention
Compliance validation against hook architecture best practices
Standardized quality scoring system with detailed reporting
Performance benchmarking for execution time and resource usage

Use Cases

Auditing third-party Claude Code plugins for security risks before deployment
Optimizing slow-performing hooks to reduce latency in AI agent workflows
Ensuring new hook implementations follow official Agent SDK callback patterns