codex-write

Orchestrate Codex CLI code generation with Claude-led reviews and fixes.

Updated Dec 17, 2025
One-click install
npx skills add https://github.com/eprouveze/HealthPulse --skill codex-write
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: codex-write
Source: https://github.com/eprouveze/HealthPulse/tree/main/.claude/skills/codex-write
Command: npx skills add https://github.com/eprouveze/HealthPulse --skill codex-write

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill automates complex code-generation workflows by orchestrating Codex CLI under Claude's guidance, reviews, and surgical fixes, enabling productive engineering without burning cycles on boilerplate.

Core Features & Use Cases

  • Codex delegation for boilerplate, large refactors, and test generation across multi-file changes.
  • Claude-led orchestration, code reviews, and surgical fixes to ensure quality.
  • Multi-LLM mode checks via .claude/llm-mode.json to determine delegation behavior.
  • Context loading from brand/configs and decision logs to improve prompt quality.

Quick Start

Describe the coding task to Claude and let Codex generate the code while Claude reviews and fixes issues.

Frequently Asked Questions about codex-write

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

FAQPage Schema
How do I automate large-scale code generation for multi-file refactors?

You can automate large-scale code generation by orchestrating Codex CLI under Claude's guidance to execute multi-file refactors, applying pre-flight checks and context loading to ensure precise prompt construction and validated changes.

What is Claude's role when delegating code generation to Codex CLI?

Claude acts as a Lead Engineer during Codex CLI code generation, performing orchestration, code reviews, and surgical fixes to validate the generated boilerplate, tests, or refactored code.

How do I generate boilerplate code across a software project without manual effort?

To generate boilerplate code without manual effort, you describe the coding task to Claude, which then loads project context, constructs precise prompts, and delegates the generation to Codex CLI before reviewing the output.

Can I use Codex CLI for automated test generation alongside refactoring?

Yes, Codex CLI supports automated test generation and multi-file refactoring within the same workflow, executing the tasks while Claude validates the changes and applies surgical fixes to ensure quality.

Does Codex CLI delegation behavior require specific configuration files?

Codex CLI delegation behavior is determined by checking a `.claude/llm-mode.json` configuration file, which dictates how Claude orchestrates the multi-LLM code generation workflow.

What are the limitations of using AI orchestration for multi-file code changes?

While AI orchestration handles boilerplate and multi-file changes, it relies on loading project context and decision logs to construct prompts, meaning poorly configured contexts may limit the accuracy of the generated code.