ai-review

Route code changes to a second AI reviewer via a CLI wrapper.

2|Updated Feb 7, 2026
One-click install
npx skills add https://github.com/eins78/agent-skills --skill ai-review-eins78
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: ai-review
Source: https://github.com/eins78/agent-skills/tree/main/skills/ai-review
Command: npx skills add https://github.com/eins78/agent-skills --skill ai-review-eins78

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires @google/gemini-cli, codex, python3, and includes scripts (resource) components.

What problem does it solve?

Automates obtaining a second AI code review from an alternative model mid-session to improve confidence and catch edge cases, especially when reviewing unfamiliar stacks.

Core Features & Use Cases

  • CLI-based wrapper to run Gemini/OpenAI reviews on diffs from the current repository
  • Supports unstaged, staged, branch, and file-specific reviews
  • Auto-context and implementation plans to provide project context to the reviewer

Quick Start

Run the Claude Code command to start an AI review of your current changes.

Frequently Asked Questions about ai-review

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

FAQPage Schema
How do I get a second AI code review on my current git diff?

To get a second AI code review, this Skill identifies code changes and routes the diff to an alternative AI reviewer via a CLI wrapper, providing mid-session quality assurance.

Can I run an AI code review on unstaged changes or specific files?

Yes, you can run an AI code review on unstaged changes, staged changes, branch diffs, or specific files across typical software projects to catch edge cases.

Do I need Python and Gemini CLI to run an AI code review mid-session?

Yes, you need Python3, the Gemini CLI, and Codex installed, as the deterministic workflow wrapper relies on these dependencies to route diffs to the alternative reviewer.

What is the best way to automate code reviews for unfamiliar tech stacks?

The best way to review unfamiliar stacks is using a CLI wrapper to route branch diffs to an alternative AI model with auto-context and implementation plans for improved confidence.

Does the AI code review wrapper enforce a timeout?

Yes, the AI code review wrapper enforces a deterministic workflow with a 1-hour timeout to ensure timely feedback and maintain guardrails during the automated review process.

How does auto-context improve AI code review quality?

Auto-context provides project context and implementation plans directly to the second AI reviewer, ensuring the generated code review understands the broader mid-session changes.