review

Analyze diffs for SQL safety, LLM trust boundaries, and conditional side effects.

4|2|Updated Mar 29, 2026
One-click install
npx skills add https://github.com/florianhorner/mammamiradio --skill review-florianhorner
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: review
Source: https://github.com/florianhorner/mammamiradio/tree/main/.claude/skills/gstack/review
Command: npx skills add https://github.com/florianhorner/mammamiradio --skill review-florianhorner

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Streamlines code review by automatically analyzing diffs for SQL safety, LLM trust boundaries, and conditional side effects, reducing risk before merging.

Core Features & Use Cases

  • Automatically detects SQL safety issues, prompt injection risks, and other structural problems in diffs.
  • Produces actionable findings with concrete evidence and recommended fixes to speed up reviews.
  • Use case: when evaluating a PR that touches data access or dynamic prompts, surface critical issues and guidance before feedback cycles.

Quick Start

Review a PR by running the review skill against the diff to obtain a prioritized list of issues and recommended changes.

Frequently Asked Questions about review

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

FAQPage Schema
How do I analyze a pull request diff for SQL safety issues?

Analyze a pull request diff for SQL safety by running the review skill against the base-branch changes to identify structural issues and return prioritized findings with recommended fixes.

What is an LLM trust boundary and how do I detect prompt injection risks in code changes?

Detect prompt injection risks in code changes by evaluating diffs to identify LLM trust boundaries, surfacing unsafe dynamic prompt interactions and returning questions to clarify ambiguous behaviors.

How do I automatically review code quality and identify conditional side effects before merging?

Review code quality and identify conditional side effects by evaluating base-branch diffs to produce actionable findings with concrete evidence, reducing merge risk before code changes ship.

Can I use automated diff review for pull requests that touch data access layers?

Yes, use automated diff review for pull requests touching data access layers to surface critical SQL safety issues and structural guidance before feedback cycles begin.

What is the best way to surface structural problems in a PR diff without manual inspection?

Surface structural problems in a PR diff without manual inspection by running an automated review that produces a prioritized list of issues, suggested fixes, and clarifying questions.

Why does my diff review return questions to clarify unsafe behaviors instead of direct fixes?

A diff review returns questions to clarify unsafe behaviors when code contains ambiguous logic or conditional side effects requiring human context, alongside the prioritized list of suggested fixes.