dev-codex-sync-settings-from-claude

name

dev-codex-sync-settings-from-claude

description

Sync the user's Claude Code workflow skills into the OpenAI Codex CLI settings repo ($HOME/.codex) as Codex-native ports, fix the Codex .gitignore for new local state, then commit and push. Use when: (1) user says '/dev-codex-sync-settings-from-claude', 'sync codex settings', 'sync claude skills to codex', 'port skills to codex', or 'update codex from claude'; (2) after updating ~/.claude workflow skills (big-plan, x, x-as-pr, x-wt-teams) and Codex should catch up; (3) the $HOME/.codex repo has drifted behind $HOME/.claude. The ports are condensed Codex-native REWRITES, never file copies.

dev-codex-sync-settings-from-claude

The user's primary agent is Claude Code, but they also run OpenAI's Codex CLI. Codex reads skills from $HOME/.codex/skills (a git repo, remote Takazudo/codex-settings). This skill re-syncs the Claude-side workflow skills into Codex whenever $HOME/.claude has moved ahead. Run it FROM Claude Code — it reads $HOME/.claude/skills and writes $HOME/.codex/skills.

The one rule that matters

Codex is NOT Claude Code. A Codex skill is a condensed, Codex-native rewrite of the Claude skill — never a copy. Claude sources run 200–1100 lines; good Codex ports run ~40–150 lines: they drop Claude-harness minutiae and translate Claude concepts (Task tool, agent teams, Opus/Sonnet/Haiku subagents) into Codex primitives (spawn_agent / send_input / wait_agent / close_agent).

Before writing or refreshing any port, read references/codex-port-contract.md — the house style and adaptation rules every port follows. Copying a Claude SKILL.md verbatim into Codex is the main failure mode; don't.

What the Codex repo tracks

$HOME/.codex/ is a git repo that intentionally tracks: