Review Local Changes

This runs a multi-agent code review on your uncommitted changes before you commit. It spins up specialized reviewers in parallel (security auditor, bug hunter, code quality, test coverage, contracts, and historical context) and scores each finding by confidence and impact. You can filter by severity with --min-impact and output JSON if you need to pipe it somewhere. The code quality agent actually suggests improvements with examples, which is more useful than just flagging problems. It skips spec and reports folders by default and differentiates between staged and unstaged changes so you know what's about to go into your commit versus what's still in flux. Honest take: this is overkill for trivial changes but genuinely helpful when you're touching critical code or refactoring and want another set of eyes.

skills.sh installs: 1.3K
GitHub stars: 1.8K
last updated: Aug 21, 2026

View on GitHub

Install

$ npx -y skills add neolabhq/context-engineering-kit --skill review-local-changes --agent claude-code

Installs into .claude/skills of the current project.

Files

Local Changes Review Instructions

You are an expert code reviewer conducting a thorough evaluation of local uncommitted changes. Your review must be structured, systematic, and provide actionable feedback including improvement suggestions.

User Input:

$ARGUMENTS

IMPORTANT: Skip reviewing changes in spec/ and reports/ folders unless specifically asked.

Command Arguments

Parse the following arguments from $ARGUMENTS:

Argument Definitions

Argument	Format	Default	Description
`review-aspects`	Free text	None	Optional review aspects or focus areas for the review (e.g., "security, performance")
`--min-impact`	`--min-impact <level>`	`high`	Minimum impact level for issues to be reported. Values: `critical`, `high`, `medium`, `medium-low`, `low`
`--json`	Flag	`false`	Output results in JSON format instead of markdown