Does this skill work on projects without Git?

Yes, for non-git projects, the analyzer falls back to identifying recently modified files to determine the appropriate scope for the review package.

Can I use this for security audits?

Yes, you can specify 'Security' as a focus area during the requirement phase to have the analyzer prioritize vulnerabilities and authentication patterns in the generated prompt.

How do I share the package with another AI model?

The skill generates a ZIP file and copies a custom instructional prompt to your clipboard. Simply upload the ZIP to the other AI and paste the prompt to start the review.

What files are included in the review package?

The skill automatically identifies core modified files, related tests, and relevant configuration files based on your recent activity, while excluding binaries and build artifacts.

Review Package Creator

Name: Review Package Creator
Author: petekp

bypetekp

•

コラボレーションツール

Bundles current code changes into a self-contained package with contextual documentation for seamless external AI or human review.

The Review Package skill automates the complex process of preparing code for external feedback by gathering relevant files, generating a contextual README, and crafting a specialized instructional prompt. It uses a dedicated subagent to analyze your codebase, identifying core changes versus supporting files to ensure the reviewer has the perfect amount of context. Whether you need a second opinion from another AI model or are preparing a hand-off for a human lead, this skill creates a portable, self-contained ZIP file that eliminates the friction of manual context-setting.

主な機能

01Generates a tailor-made instructional prompt automatically copied to your clipboard

02Targeted focus areas including Performance, Security, and Maintainability

032 GitHub stars

04Customizable review scopes for implementation, architecture, or comprehensive feedback

05Automated codebase analysis to identify core changes and related dependencies

06Self-contained ZIP creation with preserved directory structures and excluded artifacts

ユースケース

01Human hand-off: Preparing a clean, documented bundle for a senior developer or peer code review

02Contextual auditing: Creating focused snapshots of specific features for security or performance analysis

03Cross-model validation: Getting a 'second opinion' on Claude's work from another AI model like GPT-4o

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add petekp/agent-skills review-package

For use in Claude.ai and ChatGPT

Download Skill