grumpy-review

Name: grumpy-review
Availability: InStock
Author: lklimek

Run parallel-agent code reviews and produce deduplicated, severity-ranked reports.

1|Updated Feb 20, 2026

Search Tags:#code quality #developer tools #quality assurance #code review #security audit #ai assistant

Ready to run

npx skills add https://github.com/lklimek/claudius --skill grumpy-review

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: grumpy-review
Source: https://github.com/lklimek/claudius/tree/main/skills/grumpy-review
Command: npx skills add https://github.com/lklimek/claudius --skill grumpy-review

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires jsonschema, and includes scripts (resource) components.

What problem does it solve?

This Skill automates the tedious and often inconsistent process of code reviews, identifying potential issues in code quality, security, and documentation.

Core Features & Use Cases

Multi-Agent Review: Leverages specialist AI agents for security, code quality, project consistency, and documentation.
Deduplicated Reporting: Consolidates findings from multiple agents into a single, ranked report.
Interactive Triage: Allows users to classify findings (Fix, Accept Risk, Defer, False Positive) before code changes are made.
Use Case: Before merging a new feature branch, run /grumpy-review to get a comprehensive audit of potential issues, then use /triage-findings to decide how to address each one.

Quick Start

Run a comprehensive code review on the current branch by executing the command /grumpy-review.

Frequently Asked Questions about grumpy-review

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

FAQPage Schema

How do I automate a security audit and code quality review before merging a branch?▼

Run a multi-agent code review to scope changes, select specialist agents for security and quality, and output a deduplicated, severity-ranked report. This identifies potential issues in quality, security, and documentation before merging.

How does multi-agent code review work for checking dependencies and documentation?▼

Multi-agent code review works by selecting an agent mix including language specialists and technical writers to analyze code changes in parallel. A two-phase script-driven process then consolidates findings into a single report.

Can I triage and classify code review findings before making code changes?▼

You can triage code review findings by interactively classifying each issue as Fix, Accept Risk, Defer, or False Positive. This allows you to decide how to address findings before modifying your codebase.

Does the code review process require specific dependencies like jsonschema?▼

The code review process requires the jsonschema dependency to support its script components. These scripts execute the two-phase consolidation process to deduplicate findings from multiple AI agents.

What is the best way to get a deduplicated and severity-ranked code review report?▼

Get a deduplicated, severity-ranked report by executing a parallel-agent review for quality, security, dependencies, and documentation. The script-driven process consolidates findings to eliminate duplicates and rank severity.