skills.rest
figuretu
figuretuCOMMUNITY
Explore all skills by this author →
GitHub

code-review-with-codex

Run parallel Claude and Codex reviews via codex exec CLI and output a consolidated report.

16|2|Updated Feb 13, 2026
Search Tags:#codex#claude#code-review#software-engineering#ai-review#review-checklist#dual-agent
One-click install
npx skills add https://github.com/figuretu/my-skills --skill code-review-with-codex
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: code-review-with-codex
Source: https://github.com/figuretu/my-skills/tree/main/code-review-with-codex
Command: npx skills add https://github.com/figuretu/my-skills --skill code-review-with-codex

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Claude 与 Codex 双视角 Code Review。并行执行 Claude review 和 Codex review（通过 codex exec CLI），对比分析后输出综合报告到 .ai_docs/review/。适用场景：(1) 用户要求双 agent 协作 review，(2) 用户手动明确触发该 skill。codex 永远不使用本 skill。

Core Features & Use Cases

  • Parallel Claude and Codex reviews using codex exec CLI to save time and improve coverage.
  • Consolidated output to .ai_docs/review/ allowing easy sharing and traceability.
  • Suitable for dual-agent collaboration and explicit user-triggered reviews on PRs or local diffs.

Quick Start

Provide the code changes or PR details and trigger this skill to start parallel Claude and Codex reviews, then wait for the consolidated report.

Frequently Asked Questions about code-review-with-codex

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

FAQPage Schema
How does a dual-agent code review compare to a standard AI review?▼

Dual-agent code review executes Claude and Codex parallel reviews, contrasting their findings to produce a consolidated evaluation with richer coverage than a single AI agent.

How do I run a dual-agent code review on a local diff?▼

Provide local diff or pull request details to trigger the skill, which executes parallel reviews via codex exec CLI and outputs a unified report to the .ai_docs/review/ directory.

Does dual-agent code review work with the codex exec CLI?▼

Yes, the dual-agent code review explicitly uses the codex exec CLI to execute the Codex-based review alongside Claude, running both in parallel to save time and improve coverage.

What is the best way to consolidate AI code review reports for pull requests?▼

The best way to consolidate AI code review reports is using a dual-agent approach that merges Claude and Codex evaluations into a single unified file for easy sharing and traceability.

When should I not use dual-agent code review?▼

You should not use dual-agent code review when Codex is executing this skill itself, as Codex never uses this skill, or when a single-agent quick check is sufficient for your changes.

Related Skills

Explore More Software Engineering Skills→
gyy0592gyy0592
7

codex-fix

Diagnose and repair Codex CLI sandbox and stream disconnection errors.

Community
Intermediate
mindroom-aimindroom-ai
232

address-pr-review-comments

Analyze and categorize GitHub PR review comments for triage and remediation.

Official
Intermediate
tonombtonomb

code-review

Compare HEAD against a fixed point to review code standards and spec alignment.

Community
Advanced
vultukvultuk
2

review-changes

Orchestrate parallel multi-agent code reviews into a single markdown report.

Community
Advanced