security-reviewer

Trace attack paths from untrusted inputs to dangerous sinks in code diffs.

1|Updated Apr 4, 2026
One-click install
npx skills add https://github.com/ybbms777/compound-engineering --skill security-reviewer-ybbms777
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: security-reviewer
Source: https://github.com/ybbms777/compound-engineering/tree/main/skills/agent-security-reviewer
Command: npx skills add https://github.com/ybbms777/compound-engineering --skill security-reviewer-ybbms777

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill solves the critical gap of missing exploitable security vulnerabilities during standard code reviews, which can lead to costly data breaches and system compromises in production environments.

Core Features & Use Cases

  • Targeted High-Risk Focus: Automatically activates when code diffs touch authentication middleware, public API endpoints, user input handling, or permission checks to prioritize the most vulnerable areas of code.
  • Attacker-Perspective Analysis: Reviews code by tracing potential attack paths from untrusted input entry points to dangerous sinks, rather than running generic compliance checklists that miss real-world exploit paths.
  • Confidence-Calibrated Findings: Reports only actionable security issues with clear confidence ratings, filtering out low-probability theoretical attacks and generic hardening advice to reduce review noise. Use case: For example, when a developer modifies a user authentication endpoint, this skill will trace if user-supplied input is properly sanitized to prevent SQL injection, auth bypass, or privilege escalation attacks.

Quick Start

Use the security-reviewer skill to analyze the current code diff for exploitable security vulnerabilities in authentication, public endpoints, and user input handling logic.

Frequently Asked Questions about security-reviewer

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

FAQPage Schema
How do I find exploitable vulnerabilities in code diffs during security reviews?

To find exploitable vulnerabilities in code diffs, trace potential attack paths from untrusted input entry points to dangerous sinks. This method prioritizes actionable exploitable flaws over generic hardening advice in security reviews.

When do I need to trace attack paths for input sanitization in my codebase?

You need to trace attack paths for input sanitization when code changes touch authentication middleware, public API endpoints, user input processing, or permission check logic to prevent SQL injection and privilege escalation.

How does confidence-calibrated vulnerability detection work for auth security?

Confidence-calibrated vulnerability detection works by reporting only actionable security issues with clear confidence ratings. It filters out low-probability theoretical attacks to reduce review noise in auth security workflows.

What is the best way to detect auth bypass and privilege escalation in application code?

The best way to detect auth bypass and privilege escalation is reviewing code from an attacker's perspective, mapping untrusted user-supplied input to dangerous execution sinks rather than running generic compliance checklists.

Does this security review approach work for generic hardening or only specific exploit prevention?

This security review approach focuses on specific exploit prevention rather than generic hardening. It targets high-risk areas like authentication middleware and public endpoints to catch real-world exploitable paths.