review

Analyze PR diffs for SQL safety, LLM trust boundary, and side-effect risks.

Updated Mar 26, 2026
One-click install
npx skills add https://github.com/tropeks/Vitali --skill review-tropeks
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: review
Source: https://github.com/tropeks/Vitali/tree/main/.claude/skills/gstack-review
Command: npx skills add https://github.com/tropeks/Vitali --skill review-tropeks

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Diff-heavy PR reviews are error-prone and slow. This skill automates pre-merge analysis to surface SQL safety risks, LLM prompt boundary violations, and conditional side effects in diffs, helping reviewers act quickly.

Core Features & Use Cases

  • Analyzes diffs for potential SQL-safety issues and risky database operations.
  • Flags LLM trust boundary violations and potential prompt-injection vectors.
  • Highlights conditional side effects and cross-cutting impacts to prevent regressions.
  • Integrates into code-review workflows to provide proactive merge-safety guidance.

Quick Start

Provide the PR diff or a URL to the PR and invoke the review to obtain a structured risk assessment.

Frequently Asked Questions about review

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

FAQPage Schema
How do I check a pull request diff for SQL safety issues before merge?

To check a pull request diff for SQL safety issues before merge, you provide the PR diff or URL to trigger a pre-landing analysis. This surfaces risky database operations and guides reviewers with actionable mitigations.

What are LLM trust boundary violations in a code review?

LLM trust boundary violations are potential prompt-injection vectors introduced in a diff. Pre-merge analysis scans for these vulnerabilities to prevent unintended LLM behavior and ensure fast, safe pull request reviews.

How can I detect conditional side effects in a PR diff automatically?

You can detect conditional side effects in a PR diff automatically by running an automated pre-merge analysis against the base branch. This highlights cross-cutting impacts and regressions to help reviewers decide whether to merge.

Does automated pre-landing review work without integrating external components?

Yes, automated pre-landing review works without external components because the skill has no dependencies. You simply provide the PR diff or URL to obtain a structured risk assessment immediately.

What is the best way to surface cross-cutting impacts in a diff-heavy PR?

The best way to surface cross-cutting impacts in a diff-heavy PR is applying automated diff analysis against the base branch. It proactively flags conditional side effects to prevent regressions and accelerate merge-safety guidance.