Definition
Slash commands
Slash commands are actions prefixed with / that trigger built-in or custom behaviour in an AI coding agent. Claude Code includes /help, /clear, /compact, /model, /effort and /simplify; Codex includes /model, /new, /review and /status. Typing / shows the live list, which is the only reliable inventory.
Why it matters
Custom slash commands are how a team’s way of working stops living in somebody’s head. A review checklist or a release procedure becomes one word that anyone can type and every run repeats the same way.
How each tool handles it
- Cursor
- Commands depend on the current agent surface.
- Claude Code
-
/help,/clear,/compact,/model,/effort,/simplify, plus user-defined commands. There is no/think. - Codex
-
/model,/new,/review,/status, plus user-defined commands.
Frequently asked questions
Is there a `/think` command in Claude Code?
No. think is ordinary prompt text, not a command. What adjusts reasoning depth is /effort, and plan mode is entered with Shift+Tab or /plan.
How do I see which commands exist?
Type /. Both Claude Code and Codex show the live list, which is the only inventory that stays correct — the set changes between releases, so any written list ages.
Can I write my own slash commands?
Yes, in all three tools: Cursor, Claude Code and Codex all support user-defined commands alongside the built-in ones. That is how a team's review checklist or release procedure becomes one word instead of a document somebody has to remember.