cathrynlaverycathrynlaveryCommunityยท1 Agent Skills Included

codex-skill

Independent second-opinion review of coding plans and architecture

Reviews every coding plan automatically before approval by consulting OpenAI's Codex for independent verification. Catches missing steps, edge cases, and risky architecture decisions that a single model might overlook. Supports manual queries for researching unfamiliar APIs, verifying code patterns, and comparing implementation approaches. Offers fast, balanced, and deep reasoning modes to match each question's complexity.
npx skills add cathrynlavery/codex-skill --all -g -y

All Skills in This Repository (1)

Pure Emerald Level Indicators

Frequently Asked Questions

FAQPage Schema
How to install codex-skill?โ–ผ

Run `npx skills add cathrynlavery/codex-skill --all -g -y` in your terminal to install the skill and its automatic plan review hook globally.

How to get a second opinion on coding plans?โ–ผ

Once installed, every plan your coding assistant creates is automatically sent to Codex for review before you approve it, flagging risks and missing steps.

What does codex-skill require to run?โ–ผ

You need the OpenAI Codex CLI installed and a configured OpenAI API key, since the skill calls Codex for independent analysis.

Can I ask Codex questions manually?โ–ผ

Yes. Invoke the skill directly or ask in plain English to verify an approach, research a library, or get an alternative perspective on architecture.

Which Codex model does it use?โ–ผ

It defaults to the balanced gpt-5.6-terra model, escalates to gpt-5.6-sol for deep architecture analysis, and drops to gpt-5.6-luna for quick fact checks.

Related Repositories in Software Engineering

View All in Software Engineeringโ†’