lploc94Community·8 Agent Skills Included

codex_skill

Dual-reviewer code, plan, PR, and security review debates

Runs adversarial code reviews where Claude and Codex independently analyze plans, uncommitted changes, commits, pull requests, and full codebases, then debate findings until consensus. Catches bugs, security vulnerabilities, and design flaws that a single reviewer misses, without manual diff reading or copy-pasting code. Includes chunked large-codebase audits and OWASP-based security scans to ship safer code faster.

69|23

↗

npx skills add lploc94/codex_skill --all -g -y

Available:

Explains to the agent how the eight review skills are installed and orchestrated through the shared codex-runner, including session lifecycle, debate loops, and error handling rules.

All Skills in This Repository (8)

Pure Emerald Level Indicators

📦 In Repo

lploc94

codex-plan-review

Facilitate adversarial AI agent debates to review Markdown implementation plans.

codex-commit-review

Review commit messages for clarity and accuracy against code diffs using Codex CLI.

codex-parallel-review

Coordinates Claude and Codex agents to review code and reach consensus reports.

codex-codebase-review

Review entire codebases by chunking files into modules and synthesizing cross-cutting issues.

codex-security-review

Analyze code for security vulnerabilities using OWASP Top 10 and CWE patterns.

codex-impl-review

Automate adversarial code review of uncommitted changes or branch differences using the Codex CLI.

codex-think-about

Facilitate structured technical debates between Claude Code and OpenAI Codex.

codex-pr-review

Automates pull request review using Codex CLI to identify code quality issues.

Community

Advanced

Frequently Asked Questions

FAQPage Schema

How to install codex_skill?▼

Run `npx skills add lploc94/codex_skill --all -g -y` in your terminal to install all eight review skills globally.

What does codex-review do?▼

It makes Claude and OpenAI Codex independently review your plans, code changes, commits, and PRs, then debate disagreements until they reach a consensus verdict.

Can it review large codebases?▼

Yes. The codebase-review skill splits projects of 50 to 500+ files into module chunks, reviews each separately, and synthesizes cross-cutting findings.

Does it check for security vulnerabilities?▼

Yes. The security-review skill scans code against the OWASP Top 10 and CWE patterns, reporting severity, attack vectors, and suggested fixes.

What do I need to run codex_skill?▼

You need Node.js 22 or later, Claude Code, and the OpenAI Codex CLI with a configured API key available in your PATH.