agentic-security:security-eval-warn

Official

Block runtime code-eval from untrusted input.

AuthorClear-Capabilities
Version1.0.0
Installs0

System Documentation

What problem does it solve?

Prevents runtime code-eval on user input, reducing the risk of remote code execution by blocking untrusted strings from triggering eval-style calls.

Core Features & Use Cases

  • Detection: Identifies common eval and sandbox-escape patterns across languages (JS/TS: eval, new Function, setTimeout with string, vm calls; Python: eval/exec/compile/import, globals; Ruby: eval/class_eval/instance_eval; PHP: eval, assert, create_function; Shell-from-JS: exec, execSync, child_process variants; Templating engines: Mustache.render, Handlebars.compile with user input).
  • Protection: Refuses the evaluated input and proposes a structured, safer alternative.
  • Use Case: Guarding an API endpoint that accepts user-provided code snippets to ensure no runtime execution occurs.

Quick Start

Refuse any runtime code-eval request from untrusted input and propose a safe, sandboxed alternative.

Dependency Matrix

Required Modules

None required

Components

Standard package

💻 Claude Code Installation

Recommended: Let Claude install automatically. Simply copy and paste the text below to Claude Code.

Please help me install this Skill:
Name: agentic-security:security-eval-warn
Download link: https://github.com/Clear-Capabilities/agentic-security/archive/main.zip#agentic-security-security-eval-warn

Please download this .zip file, extract it, and install it in the .claude/skills/ directory.
View Source Repository

Agent Skills Search Helper

Install a tiny helper to your Agent, search and equip skill from 510,000+ vetted skills library on demand.