Skip to content

Slash commands — human-readable reference

Cursor lets a workspace define slash commands: you type /something in chat and the assistant follows a short recipe. Commands are optional and only appear when the project includes .cursor/commands/ (often in a private or team template repo).

This page summarizes common commands from the where_cursor_feels_home-style workspace so the onboarding cheatsheet works on a bare clone of ai-assisted-work (no .cursor/ folder here).


Commands (summary)

Command When to use What it typically does
/journal End of day or after a meaningful session Create or update today’s journal; may draft Backlog candidates and Glossary candidates.
/session-end You are closing a session Wrap-up: journal sections, backlog/glossary candidates, update activeContext.md, optional session-log.md.
/pm Product thinking Load product management context — discovery, prioritization, roadmaps, AI product risks; reads product-management/PM_KNOWLEDGE.md when present.
/pmo Delivery / process PMO context — Agile, Lean, Six Sigma, CRISP-DM; reads pmo/PMO_KNOWLEDGE.md when present (not in minimal public tree).
/capture-idea Mid-task “park this thought” Append a quick line to BACKLOG_INBOX.md (or equivalent) without derailing the current build — only if that file exists.
/canine Dog / puppy questions Behavior and welfare guidance when a dog-agent (or similar) stack is in the repo.
/collaboration-check Things feel off with the assistant Short collaboration health check between you and the agent.

If you do not have these files

You can still ask in plain language: e.g. “Help me wrap up this session: draft journal bullets and open questions from this chat.” See cursor-triggers-and-prompts.md.


Full workspace–only pieces

Backlog inbox, transcript scripts, HTML handbook sync, and skill files are documented in ../INDEX-full-workspace-only.md.