Codex CLI Bridge (Claude Code ↔ Codex)

What it does

Creates a cross-tool bridge between Claude Code projects and the OpenAI Codex CLI. It can parse CLAUDE.md, generate a reference-based AGENTS.md (no duplication), and expose Python execution helpers that wrap codex exec for analysis, edits, and session resume. The skill also documents how to invoke prompt-only or functional skills from the Codex CLI.

When to use it

Use when you need to make Claude Code skills consumable from Codex CLI workflows, to generate AGENTS.md for a project, or to run Codex-based analysis/edit tasks safely (the skill includes safety checks for Codex installation and auth). Useful for teams that use both Claude Code and Codex CLI.

What's included

• Scripts: Python helpers (agents_md_generator.py, codex_executor.py, claude_parser.py, project_analyzer.py, safety_mechanism.py) — see repository files.
• References: README, examples, AGENTS.md templates.
• Instructions: AGENTS.md generation flow, codex exec usage patterns, sandbox selection guidance, and troubleshooting steps.

Compatible agents

Designed for cross-tool automation; useful to developers and coding agents that can run Python wrappers and invoke Codex CLI (gpt-5, gpt-5-codex compatible workflows).