Dialect cheat sheet
The same primitive in both tools, side by side — grouped by concept area, each row linking to the lesson that teaches it. This is the quick-scan view of the feature-parity matrix; for tier nuance, translation guides, and dated version claims, follow a row through to the matrix. As of .
Lead with
Getting started
| Concept | Claude Code | Codex | Lesson |
|---|---|---|---|
| install & auth T1 | install + auth, approval prompts on first tool use | install + auth, approval prompts on first tool use | A1 |
| built-in slash commands T1 | /model /clear /compact /rewind /permissions | /model /approvals /skills /agent | A2 |
| frontier models T1 | Claude (Fable / Opus / Sonnet / Haiku tiers) + effort levels | GPT-5.3-Codex / 5.4 / 5.5 + effort levels | A2 |
Instruction files & memory
| Concept | Claude Code | Codex | Lesson |
|---|---|---|---|
| instruction files T1 | CLAUDE.md (project + user scopes) · auto-memory · path-scoped .claude/rules/ | AGENTS.md hierarchy (global ~/.codex/, repo root, nested dirs) · AGENTS.override.md · 32 KiB budget | B1 |
| auto-memory T3 · CC | auto-memory — the agent updates CLAUDE.md from the conversation | nearest equivalent: deliberate AGENTS.md updates (you / the agent edit it explicitly) | B1 |
Settings & safety
Skills
| Concept | Claude Code | Codex | Lesson |
|---|---|---|---|
| skills T1 | .claude/skills/<name>/SKILL.md — invoke /name | .agents/skills/<name>/SKILL.md — invoke $name (agentskills.io standard, $skill-creator) | C1 |
Hooks
| Concept | Claude Code | Codex | Lesson |
|---|---|---|---|
| hooks T1 | hooks block in settings.json — PreToolUse, PostToolUse, SessionStart, SubagentStop, Stop, … | hooks.json or [hooks] in config.toml (stable v0.124+) — same event vocabulary | C2 |
| hook events T1 | PreToolUse, PostToolUse, SessionStart, SubagentStop, Stop, … | same vocabulary: PreToolUse, PostToolUse, SessionStart, SubagentStop, Stop, PreCompact, … | C2 |
| agent-based hook gates T3 · CC | prompt / agent-based hooks — an LLM-judge gate inside the hook | nearest equivalent: command gates + Automatic Review Agent | C2 |
MCP & data access
| Concept | Claude Code | Codex | Lesson |
|---|---|---|---|
| MCP registration T1 | .mcp.json / claude mcp add | [mcp_servers] in config.toml / codex mcp add | C3 |
| see a UI / web page T2 | Playwright MCP screenshots | computer use · in-app browser · Appshots | C3 |
| computer use / Appshots T3 · CX | nearest equivalent: Playwright MCP (drive + screenshot a browser) | in-app browser · computer use · Appshots (native GUI control) | C3 |
Subagents & orchestration
| Concept | Claude Code | Codex | Lesson |
|---|---|---|---|
| subagent definitions T1 | .claude/agents/<name>.md (markdown + frontmatter) · built-ins Explore / Plan / general-purpose | .codex/agents/<name>.toml (TOML) · built-ins default / worker / explorer | D1 |
| concurrency control T1 | per-workflow cap (≤16) | [agents] max_threads (default 6) | D4 |
| NotebookEdit + chart outputs T3 · CC | NotebookEdit — edits .ipynb cells directly and reads chart outputs | nearest equivalent: jupytext / script workflow + image input | D1 |
Sessions, planning & parallel work
| Concept | Claude Code | Codex | Lesson |
|---|---|---|---|
| sessions, resume & fork T2 | sessions / resume / fork · checkpoints + /rewind | conversation forking + git · codex exec resume | B2 |
| Plan mode T3 · CC | Plan mode + built-in Explore / Plan subagents | nearest equivalent: read-only approval profile + explorer subagent | B2 |
| parallel-agent isolation T2 | git worktrees + worktree-spawned agents (composable local primitive) | desktop-app parallel threads (native worktrees) + isolated cloud tasks (managed delegation) | D2 |
Long-running & scheduled work
| Concept | Claude Code | Codex | Lesson |
|---|---|---|---|
| long-running supervision T2 | /loop (recurring self-paced prompt) + background agents | Goal Mode (objective-driven multi-hour run) | D3 F1 |
| massively parallel fan-out T2 | Dynamic Workflows (pipeline() JS orchestration) + ultracode | spawn_agents_on_csv + cloud best-of-N | D4 |
| spawn_agents_on_csv T3 · CX | nearest equivalent: Dynamic Workflow pipeline() over a manifest | spawn_agents_on_csv (experimental) — one agent per CSV row, cloud best-of-N | D4 |
| scheduled autonomous work T2 | Routines (cloud; cron / GitHub / API triggers) | Codex cloud tasks + GitHub integration | E2 |
| automatic-research orchestration T2 | /loop orchestration over subagents + worktrees + headless reruns (compose A–E locally) | Goal Mode + cloud tasks + best-of-N (managed cloud orchestration) | F1 |
Headless, SDK & CI
| Concept | Claude Code | Codex | Lesson |
|---|---|---|---|
| headless run T1 | claude -p "..." --output-format json | codex exec "..." --json | E1 |
| SDK T1 | Agent SDK (Python + TypeScript) | Codex SDK (TypeScript) | E1 |
| CI integration T1 | GitHub Actions (claude-code-action) | Codex GitHub Action · @codex on PRs/issues | E1 |
| automatic PR review T3 · CX | nearest equivalent: a review subagent / agent-based gate inside the GitHub Action (claude-code-action) | @codex on a PR/issue · the Automatic Review Agent reviews every PR | E1 |
Distribution & frontier
| Concept | Claude Code | Codex | Lesson |
|---|---|---|---|
| plugins & marketplaces T1 | plugins + marketplaces | plugins + marketplaces (agentskills.io publishing) | E3 |
| Linear / app integrations T3 · CX | nearest equivalent: MCP servers (connect the app via MCP) | first-party Linear + app integrations | E2 |
| output styles T3 · CC | output styles — switch the agent's writing voice/format | nearest equivalent: style directives in AGENTS.md | F1 |
| agent teams T3 · CC | agent teams (experimental) — coordinated multi-agent roles | nearest equivalent: multiple cloud tasks + shared repo | F1 |
Command quick-ref by unit
The verbs you reach for at each stage of the expedition. Units A–D are live; E–F ship as their lessons land — listed here so the field log is complete.
| Unit | Claude Code | Codex |
|---|---|---|
| A Everyday | claude · /model · /clear · /compact · /rewind · /permissions | codex · /model · /approvals · /skills · /agent |
| B Memory & safety | CLAUDE.md · .claude/rules/ · .claude/settings.json · plan mode | AGENTS.md · ~/.codex/config.toml · profiles · read-only approval |
| C Skills · hooks · MCP | /skill-name · hooks{} in settings.json · claude mcp add | $skill-name · hooks.json / [hooks] · codex mcp add |
| D Orchestration | .claude/agents/*.md · /loop · Dynamic Workflows · git worktree | .codex/agents/*.toml · Goal Mode · spawn_agents_on_csv · cloud tasks |
| E Automation | claude -p --output-format json · Routines · claude-code-action | codex exec --json · cloud tasks · @codex · Codex GitHub Action |
| F Automatic research | /loop orchestration · output styles · agent teams (exp.) · cost/effort tuning | Goal Mode loop · AGENTS.md style directives · multiple cloud tasks · effort tuning |
Tier tags: T1 parity · T2 same job, different primitive · T3 one-sided (owner noted). Every claim here is dated and revisited quarterly — the parity matrix carries the translation guides and watch-this-space notes.