review-safe

Audit codebases read-only for security and correctness issues.

Updated Mar 17, 2026

Search Tags:#dependencies #static-analysis #code-review #read-only #security-audit #findings #error-detection

Ready to run

npx skills add https://github.com/Rheinmir/skills-kit --skill review-safe

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: review-safe
Source: https://github.com/Rheinmir/skills-kit/tree/main/skills/review-safe
Command: npx skills add https://github.com/Rheinmir/skills-kit --skill review-safe

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps security-minded teams identify and report issues in codebases without modifying any code, saving time and reducing risk through structured findings.

Core Features & Use Cases

Scope-wide checks for security, logic, correctness, and dependency analysis across target files.
Structured findings with a consistent format for easy triage.
Non-intrusive review runs in read-only mode, no edits or fixes.

Quick Start

Review the target codebase in read-only mode and generate findings only.

Frequently Asked Questions about review-safe

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

FAQPage Schema

How do I run a read-only code audit to identify security and correctness issues?▼

Perform a read-only code audit by scanning target repositories or diffs in parallel for security, logic, correctness, and dependency issues to generate standardized, actionable findings without modifying any code.

What is the best way to perform static analysis on code diffs without modifying the repository?▼

Static analysis on code diffs runs in a read-only mode to detect security and correctness issues across changes, producing structured findings that prioritize actionable issues while avoiding any code edits or fixes.

Can I audit dependency vulnerabilities across an entire repository without making changes?▼

Yes, you can audit dependency vulnerabilities across an entire repository because the review process operates strictly in read-only mode, performing parallel checks and generating findings without applying fixes.

Does a non-intrusive security audit produce structured findings for triage?▼

A non-intrusive security audit produces structured findings with a consistent format, allowing teams to easily triage identified security, logic, and correctness issues without introducing risk through code modifications.

What limitations exist when using a read-only code review for error detection?▼

The primary limitation of this read-only code review is that it avoids fixes or edits entirely, meaning users receive actionable findings and must implement the actual code corrections manually.

How do I check for logic and correctness errors in a codebase using static analysis?▼

Check for logic and correctness errors by applying static analysis across target files in parallel, which identifies issues and outputs standardized findings that highlight actionable problems for review.