kaizen-review-pr

Automate adversarial PR review across dimensions with a multi-phase workflow.

Updated Mar 16, 2026
One-click install
npx skills add https://github.com/Garsson-io/kaizen --skill kaizen-review-pr
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: kaizen-review-pr
Source: https://github.com/Garsson-io/kaizen/tree/main/.agents/skills/kaizen-review-pr
Command: npx skills add https://github.com/Garsson-io/kaizen --skill kaizen-review-pr

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Complex PR reviews across multiple dimensions are slow and error-prone when done manually; this Skill automates adversarial checks and enforces a robust fix-loop.

Core Features & Use Cases

  • Automated adversarial checks across dimensions during PR reviews
  • Parallelized dimension analysis with consolidated findings
  • Per-dimension storage and a structured fix loop with escalation

Quick Start

Provide a PR URL to start a review round and let Kaizen surface adversarial dimension findings.

Frequently Asked Questions about kaizen-review-pr

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

FAQPage Schema
How do I automate adversarial PR code reviews across multiple dimensions?

Automated PR review coordinates pre-fetching diff, body, and issue data to run parallel dimension analyses, classify findings, and enforce a fix loop. This surfaces actionable code review findings across dimensions without manual checks.

What is the best way to structure a fix loop during an automated PR review?

A structured PR review fix loop uses a multi-phase workflow: Run Dimensions, Classify, Fix, and Verdict. It stores findings per-dimension and applies strict review gates with escalation to ensure all adversarial checks pass before merging.

How does parallel dimension analysis work when reviewing a pull request?

Parallel dimension analysis evaluates different aspects of a PR simultaneously by coordinating pre-fetched diff, PR body, and briefing data. This parallelized review workflow consolidates findings efficiently, reducing the error rate of manual sequential checks.

Can I start an automated code review by just providing a PR URL?

Yes, providing a PR URL starts an automated review round. The workflow pre-fetches the required diff, body, and issues data to drive the adversarial dimension analyses and surface actionable findings without manual configuration.

When do I need per-dimension findings storage in a code review workflow?

Per-dimension findings storage is needed when running complex, parallel PR review analyses. It isolates specific check results within the fix loop, enabling strict review gates and structured escalation before a final verdict is reached.