Which models does Code with Codex use?

It automatically selects from the GPT-5.1 and GPT-5.2 Codex series based on the complexity level, optimizing for both cost and output quality.

What is the primary purpose of Code with Codex?

It is a specialized skill for Claude Code that leverages memex-cli to handle advanced coding tasks like project scaffolding, refactoring, and complex algorithm implementation through a structured execution protocol.

What is task decomposition in this context?

For complex tasks (L3+), the skill automatically breaks down the project into independent subtasks, assigns IDs, and establishes dependencies to ensure logical and parallel execution.

How does the Complexity Assessment (L1-L5) work?

The skill uses a decision tree to evaluate your request, ranging from L1 (single file scripts) to L5 (multi-module microservices), ensuring the most efficient model and strategy are used.

Does this skill support parallel execution?

Yes, for L2 and higher tasks, the skill analyzes dependencies to build a Directed Acyclic Graph (DAG), allowing independent modules to be generated simultaneously.

Code with Codex

Name: Code with Codex
Author: chaorenex1

bychaorenex1

•

API開発

Automates complex code generation and architectural tasks using a structured multi-level execution protocol and task decomposition.

概要

Code with Codex is a sophisticated Claude Code skill designed to handle everything from simple scripts to complex microservice architectures. By leveraging the memex-cli and a Codex-powered backend, it implements a mandatory execution protocol that assesses task complexity (L1-L5), performs dependency analysis, and decomposes large projects into manageable subtasks. This skill is essential for developers who need to generate high-quality, production-ready code with memory support, parallel execution, and automated model selection optimized for specific task requirements.

主な機能

Intelligent Model Selection (GPT-5 Codex series)
Multi-level Complexity Assessment (L1-L5) with auto-grading
Parallel execution of independent subtasks for speed
Advanced Dependency Analysis and DAG execution mapping
Automated Task Decomposition for large-scale projects
1 GitHub stars

ユースケース

Creating comprehensive unit and integration test suites in parallel
Refactoring large codebases with automated dependency mapping
Generating complex multi-file project scaffolds and microservices

概要

主な機能

Intelligent Model Selection (GPT-5 Codex series)
Multi-level Complexity Assessment (L1-L5) with auto-grading
Parallel execution of independent subtasks for speed
Advanced Dependency Analysis and DAG execution mapping
Automated Task Decomposition for large-scale projects
1 GitHub stars

ユースケース

Creating comprehensive unit and integration test suites in parallel
Refactoring large codebases with automated dependency mapping
Generating complex multi-file project scaffolds and microservices