review

Detect pre-landing risks in a pull request diff against its base branch.

6|2|Updated Oct 18, 2021
One-click install
npx skills add https://github.com/jerrywu001/jerrywu001 --skill review-jerrywu001
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: review
Source: https://github.com/jerrywu001/jerrywu001/tree/main/.claude/skills/review
Command: npx skills add https://github.com/jerrywu001/jerrywu001 --skill review-jerrywu001

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Pre-landing PR review helps prevent risky code from being merged by scanning a diff against the base branch for SQL safety, LLM trust-boundary violations, conditional side effects, and other structural issues that tests might miss.

Core Features & Use Cases

  • SQL safety checks: flags unsafe query construction patterns and likely injection risks when changes touch database code.
  • LLM trust boundary review: detects places where LLM output could be treated as trustworthy input without proper validation or containment.
  • Structural issue detection: surfaces race conditions, concurrency hazards, enum/value completeness gaps, and conditional side effects that can behave differently in production.
  • Use case: before you merge a feature branch, run a pre-landing review to ensure the diff is safe enough to ship, especially when it touches persistence, prompts, or control-flow logic.

Quick Start

Ask the AI to run the review on your PR diff by saying: Review this PR for SQL safety, LLM trust boundary issues, conditional side effects, and structural risks.

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 injection and race conditions before merging?

Pre-landing code review scans your pull request diff against the base branch to detect SQL injection risks, race conditions, and structural hazards. It applies an ordered checklist-driven review to flag risky code before it ships.

What is an LLM trust boundary violation and how do I identify it in my code?

An LLM trust boundary violation occurs when untrusted LLM output is processed without validation. Pre-landing review detects places where LLM output could be treated as trustworthy input without proper containment or sanitization.

How do I review a feature branch for conditional side effects and incomplete enum handling?

You can run a pre-landing PR review to surface conditional side effects and enum/value completeness gaps. It analyzes the diff to catch control-flow logic and value handling issues that might behave differently in production.

Does pre-landing code review work without running my test suite?

Yes, pre-landing code review works independently of tests by analyzing the diff directly. It catches structural issues like SQL safety and LLM trust boundary violations that standard tests might miss.

What is the best way to automate risk detection for database code changes in a pull request?

Running a checklist-driven pre-landing review is an effective way to detect risks in database code changes. It flags unsafe query construction patterns and likely SQL injection risks directly from the diff.

When should I escalate a flagged issue during a pre-landing code review?

You should escalate during a pre-landing code review when verification or user input is needed to confirm a flagged risk. The review prioritizes fix-first actions and requires clear escalation for uncertain structural hazards.