Jelly Codex Skill
Integrates the Codex CLI to solve complex coding challenges and architectural designs using high-reasoning models like GPT-5.1.
The source repository doesn't declare a license. Check its terms before reusing the code.
Key features
- High-reasoning model execution using GPT-5.1 and GPT-5.1-codex
- Integrated web search capability for real-time information retrieval
- Non-interactive workspace-write access for automated code modifications
- Persistent session resumption to maintain context across tasks
- Advanced architectural design and implementation guidance
Use cases
- Solving advanced algorithmic challenges that require specialized reasoning models
- Designing complex system architectures and data structures with high reasoning
- Automating multi-file code refactoring and feature implementation
FAQ
How does this skill improve my developer workflow?
It streamlines the development process by providing non-interactive 'workspace-write' access for automated modifications. This means Claude can autonomously refactor files and implement features using high-tier reasoning models without requiring constant manual oversight for every step.
When should I use this skill?
You should use this skill for tasks that require advanced reasoning beyond standard LLM capabilities, such as designing complex data structures, performing deep code reviews, implementing intricate algorithms, or when you need to resume a persistent coding session with full context.
What does the Jelly Codex Skill do?
The Jelly Codex Skill integrates the Codex CLI into Claude Code, allowing users to leverage high-reasoning models like GPT-5.1 and GPT-5.1-codex. It enables Claude to handle complex architectural designs, perform automated code edits, and access real-time information through integrated web search.
Does this skill support session persistence?
Yes. The skill includes built-in logic to detect continuation requests. By using the 'codex resume' command, it can pick up exactly where a previous session left off, maintaining all previous context and architectural decisions for iterative development.
Why is 'codex exec' used instead of the standard 'codex' command?
Claude Code's bash environment is non-interactive (non-TTY). The standard 'codex' command requires a terminal to function, whereas 'codex exec' is specifically designed for automated environments, ensuring that tasks complete successfully without 'stdout' errors.
Related skills
Claude Plugin Configuration Manager
Standardizes per-project plugin settings and state management using YAML frontmatter and Markdown files within the .claude directory.
Developer Tools142,11812 ptsClaude Code Agent Architect
Guides developers through the creation, configuration, and optimization of autonomous subagents for Claude Code plugins.
Developer Tools142,11812 ptsHook Development
Create event-driven automation scripts (hooks) to validate operations, enforce policies, and customize Claude Code workflows.
Developer Tools46,24612 pts