agent-security-review

Name: agent-security-review
Availability: InStock
Author: MindFort

Reviews AI agents and LLM systems for security vulnerabilities and risks.

16|2|Updated May 26, 2026

Search Tags:#vulnerability assessment #AI security #security review #LLM security #agent auditing

Ready to run

npx skills add https://github.com/mindfortai/security-skills --skill agent-security-review-mindfortai

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: agent-security-review
Source: https://github.com/mindfortai/security-skills/tree/main/skills/agent-security-review
Command: npx skills add https://github.com/mindfortai/security-skills --skill agent-security-review-mindfortai

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill ensures the security of AI agents and LLM systems by thoroughly reviewing and analyzing them for potential vulnerabilities and risks.

Core Features & Use Cases

Comprehensive Review: Inspects for prompt injection, unsafe tool use, context leakage, approval bypasses, and missing security tests.
Auditing AI Agents: Ideal for auditing AI agents, MCP servers, tool routers, retrieval systems, memory layers, browser automation, shell/code execution, and human-approval gates.
Use Case: When you need to ensure that your AI agents are secure and not vulnerable to unauthorized actions or data leaks.

Quick Start

Run the agent-security-review skill to audit your AI agent for potential security issues.

Frequently Asked Questions about agent-security-review

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I audit AI agents for prompt injection and unsafe tool use?▼

You can audit AI agents for prompt injection and unsafe tool use by running a security review that analyzes tool use, context management, and security policies to identify vulnerabilities like context leakage and approval bypasses.

What is LLM security auditing and how does it work for MCP servers?▼

LLM security auditing is the process of reviewing AI agents and LLM systems for vulnerabilities. It works by inspecting MCP servers, tool routers, and memory layers for missing security tests, context leakage, and approval bypasses.

Can I use this to review browser automation and shell code execution agents?▼

Yes, you can review browser automation and shell code execution agents. The security review applies to auditing tool routers, retrieval systems, memory layers, and human-approval gates for potential security vulnerabilities.

What's the best way to prevent context leakage in retrieval systems?▼

The best way to prevent context leakage in retrieval systems is to perform a thorough agent security review that analyzes context management and security policies to detect unauthorized data leaks and missing security tests.

How do I check my AI agent for human-approval bypass vulnerabilities?▼

To check your AI agent for human-approval bypass vulnerabilities, run a security review that analyzes approval gates, tool use, and context management to ensure unauthorized actions cannot bypass required human confirmations.