code-review

Review pull requests for logic, architecture, performance, and security issues.

3|1|Updated Apr 12, 2026
One-click install
npx skills add https://github.com/Cogni-AI-OU/cogni-ai-agent-skills --skill code-review-cogni-ai-ou
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: code-review
Source: https://github.com/Cogni-AI-OU/cogni-ai-agent-skills/tree/main/code-review
Command: npx skills add https://github.com/Cogni-AI-OU/cogni-ai-agent-skills --skill code-review-cogni-ai-ou

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It helps prevent regressions and security vulnerabilities by guiding expert, review-only analysis of code changes and pull requests before they merge.

Core Features & Use Cases

  • Review Tests First: Prioritizes reading tests to understand intended behavior and coverage gaps.
  • Multi-Dimension Code Assessment: Evaluates correctness, robustness, readability, architecture, security (including secret/leak checks), and performance risks.
  • Actionable, Prioritized Feedback: Produces issue comments labeled with clear priorities and requires a concrete resolution path for every issue.
  • Structured Review Output: Summarizes findings with a verdict (APPROVE or REQUEST CHANGES) and a verification-story checklist.
  • Guides Review Guardrails: Enforces review-only mode and avoids over-scaffolding or ungrounded “vague complexity” critiques.

Quick Start

Use the code-review skill to review the proposed changes in the current Pull Request and produce prioritized, structured feedback with an APPROVE or REQUEST CHANGES verdict.

Frequently Asked Questions about code-review

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

FAQPage Schema
How do I perform a static code review to identify security vulnerabilities before a pull request merge?

Static code review evaluates proposed pull request changes to identify security vulnerabilities, logical flaws, and architectural drift before merge. It requires review-first test analysis to validate correctness, maintainability, and zero-trust safety.

What does reviewing tests first reveal about code changes and coverage gaps?

Reviewing tests first reveals the intended behavior and coverage gaps of proposed code changes. This approach prioritizes reading tests to understand expected logic before assessing robustness, architecture, and performance risks.

Can I use automated static analysis to check for performance bottlenecks and secret leaks in code snippets?

Automated static analysis checks code snippets for performance bottlenecks, secret leaks, and architectural drift. It assesses correctness and security by applying adversarial self-inquiry to produce actionable, prioritized feedback.

How do I generate structured feedback with an approval verdict for a pull request?

Structured feedback generation produces prioritized issue comments with clear resolution paths and a verification story. It concludes with a structured verdict to either APPROVE or REQUEST CHANGES for the pull request.

Are there limitations to review-only mode when evaluating architectural drift?

Review-only mode limitations restrict over-scaffolding and ungrounded vague complexity critiques during architectural drift evaluation. It enforces analyzing correctness and robustness without modifying the proposed code changes directly.

What's the best way to catch logical flaws in proposed code changes before deployment?

The best way to catch logical flaws is expert static code review of proposed changes before deployment. It applies adversarial self-inquiry and multi-dimension assessment to ensure correctness, maintainability, and zero-trust safety.