security-review

Name: security-review
Availability: InStock
Author: notque

Scan git diffs for vulnerabilities and output a BLOCK/FIX/APPROVE verdict.

415|44|Updated Mar 18, 2026

Search Tags:#code review #security review #vulnerability scanning #git diff #LLM-based audit #secure commits #threat reduction

Ready to run

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

💬 Prompt for ChatGPT / Claude Chat▼

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

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires python3, scripts/security-review-scan.py, Task, Agent, Grep, Glob, Bash, and includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill prevents insecure or broken code from slipping into a commit by running a two-layer security review that combines deterministic scanning with an LLM-depth inspection of the actual diff.

Core Features & Use Cases

Deterministic security scan of changed files for known vulnerability classes to catch obvious issues early and consistently.
LLM-depth security review using a structured taxonomy to identify deeper problems like authorization gaps, injection via data flow, and missing control reasoning.
Single clear verdict (BLOCK / FIX / APPROVE) with a structured report and per-finding severity.

Quick Start

Run the security review on your current git changes by telling your agent to review your changes for security.

Frequently Asked Questions about security-review

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

FAQPage Schema

How do I run a security review on my git diff before committing?▼

To run a security review on your git diff, instruct your agent to review your changes. The Skill scans the staged or working diff using a local script and an LLM-depth audit, outputting a BLOCK, FIX, or APPROVE verdict without network calls.

What is an LLM-based code review for vulnerability scanning?▼

An LLM-based code review for vulnerability scanning uses artificial intelligence to inspect git diffs for deeper security issues like authorization gaps and injection risks. This Skill merges LLM findings with deterministic scan results to provide a unified security verdict.

Does this security review tool require API keys or network access to scan changes?▼

No, this security review tool does not require API keys or network access. It executes a deterministic vulnerability scan and LLM-depth audit entirely locally over your exact git diff, ensuring secure commits without external dependencies.

Can I use this vulnerability scanner for pull and push workflows?▼

Yes, you can use this vulnerability scanner for pull and push workflows. It provides evidence-based review of changed files before committing or merging, making it suitable for securing continuous integration and branch protection processes.

What is the best way to audit changed files for injection and authorization gaps?▼

The best way to audit changed files for injection and authorization gaps is combining deterministic scanning with an LLM-depth review. This Skill loads a security taxonomy to identify complex data flow problems and outputs per-finding severity levels.

Why does the code review output a BLOCK, FIX, or APPROVE verdict?▼

The code review outputs a BLOCK, FIX, or APPROVE verdict to provide a single clear action signal for threat reduction. It merges deduplicated findings from the deterministic scan and LLM audit into a structured report for immediate commit decisions.