26 Claude shortcuts I keep missing

I've been using Claude almost daily for months and I still catch myself writing out a five-sentence instruction when a two-character shortcut already does the job.

Last week I typed "can you compress this conversation but keep the part about the migration schema and drop the tangent about CSS." Then I scrolled the slash menu by accident and saw /compact does exactly that. With instructions. In two characters.

A creator on LinkedIn pulled every Claude slash command into one clean reference. Twenty-six of them, organized so you can actually find what you need. I was a little embarrassed at how many I'd been working around.

Here's the full breakdown, grouped so you can install them once and stop typing out instructions Claude already understands.

In partnership with

Voice dictation that doesn't mangle your syntax.

Most dictation tools choke on technical language. Wispr Flow doesn't. It understands code syntax, framework names, and developer jargon — so you can dictate directly into your IDE and send without fixing.

Use it everywhere: Cursor, VS Code, Warp, Slack, Linear, Notion, your browser. Flow sits at the system level, so there's nothing to install per app. Tap and talk.

Developers use Flow to write documentation 4x faster, give coding agents richer context, and respond to Slack without breaking focus. 89% of messages go out with zero edits. Free on Mac, Windows, and iPhone.

Try Wispr Flow free

Group 1: session and conversation control

The ones you'll reach for most. Half the battle with Claude is managing context, and this group handles it cleanly.

1. /clear wipes history and resets context. Use it when you switch topics so old context stops bleeding into the new one.

2. /cost shows token usage for the current session. Quick spend check without leaving the chat.

3. /compact [instructions] compresses the conversation and lets you tell Claude what to keep. The instructions part is the hidden gem.

4. /resume [session] reopens a previous conversation by ID or name. No more scrolling history for that one thread.

5. /branch [name] spins up a new branch from the current chat. Great for exploring a tangent without losing your main thread.

6. /rewind rolls the conversation and code back to an earlier point. Lifesaver when an experiment goes sideways.

7. /rename [name] renames the current session. Leave it blank and Claude auto-generates one.

8. /export [filename] saves the conversation as plain text. Handy for sharing a session or archiving a useful exchange.

Group 2: model, usage, and plan controls

The commands a lot of people miss entirely. Model switching mid-conversation alone is worth the read.

9. /model [model] switches between Sonnet, Opus, or Haiku without leaving the chat. Haiku for quick lookups, Sonnet for most work, Opus for the heavy lift.

10. /usage shows plan limits and rate-limit status. Check before kicking off a big batch of prompts.

11. /extra-usage turns on extra usage when you hit standard limits. Crucial on crunch days.

Group 3: project setup and memory

If you live in Claude Code or run Claude across projects, this group is where things get powerful.

12. /init sets up your project with a CLAUDE.md guide. Foundation step for any serious project work.

13. /memory edits CLAUDE.md memory files and controls auto-memory. This is where you teach Claude your preferences, tools, and patterns.

14. /add-dir <path> adds a working directory so Claude can read more files during the session.

Your docs are being read by AI. Is yours ready?

Over 50% of traffic across Mintlify's customer base is now AI agents, not humans. If your docs aren't structured for agents, your product is invisible to AI. Mintlify just raised a $45M Series B to build the knowledge layer for the agent era.

Make Your Docs Agent-Ready

*Ad

Group 4: code review and security

For anyone shipping code, these three take real busywork off your plate.

15. /diff opens a diff viewer for uncommitted changes and per-turn diffs. Way faster than alt-tabbing to the terminal.

16. /security-review scans pending changes for security issues. This one alone justifies bookmarking the list.

17. /plan [description] drops Claude into plan mode, optionally with a starter task. Forces a thought-out approach before code gets written.

Group 5: permissions, agents, and skills

The configuration layer. Most folks don't realize how much customization is baked into Claude.

18. /permissions manages allow, ask, and deny rules for tool access. Set it once, stop getting prompted on every action.

19. /agents manages agent and sub-agent configurations. The doorway to multi-agent workflows.

20. /skills lists every skill available, built-in and custom. Way easier than digging through folders.

Group 6: plugins and connections

Integration commands. They keep your environment clean without forcing a full restart.

21. /plugin manages your Claude Code plugins.

22. /reload-plugins refreshes active plugins without restarting the session. Saves real time when you're iterating on a plugin config.

23. /mcp manages MCP server connections and OAuth authentication. The control room for everything you've connected to Claude.

Group 7: personalization

Small touches, but the ones that make Claude feel like yours.

24. /config opens settings for theme, model, output style, and preferences.

25. /theme changes the color theme, including light, dark, and accessibility-friendly options.

26. /color [color] sets the prompt bar color for the current session. A tiny thing, but useful when you run several sessions side by side.

The three I reach for every single day

Twenty-six is too many to install at once. Three move the needle harder than the rest.

/compact with instructions. I run this whenever a chat passes the 20-message mark. "Compress this. Keep the schema decisions and the file paths. Drop the brainstorming." Cleaner context, faster responses, less re-explaining.

/model. Most people pick one model and ride it. Knowing you can flip to Haiku for quick lookups and back to Opus for the deep work stretches your plan further than any upgrade button.

/security-review. If you ship code, this is a one-command audit before you commit. Nothing else on the list gives you that much safety for that little effort.

If you only install three, install these.

How to actually use them

Don't try to memorize twenty-six. The trick is to know that the shortcut exists, so you reach for it instead of typing five sentences.

Pin the list somewhere you'll see it. A notes app, the side of your monitor, a sticky inside your code editor. The next time you're 40 turns deep into a conversation and need to branch off without losing the thread, your brain will surface the right slash instead of asking Claude to do it longhand.

The point isn't the cheat sheet. The point is the muscle memory.

My honest take

The thing about slash commands is they're invisible until you know they exist. You can use Claude for six months and never hit one. They don't show up in the marketing. They're not in the homepage demo.

But every long instruction you've been typing out was probably already a slash command. Two characters versus two paragraphs is the difference between flow and friction. And friction is what burns plans, kills sessions, and makes Claude feel slower than it actually is.

Pick three from the list. Use them for a week. The next time you catch yourself typing out a paragraph, your hand will reach for the slash instead.

Check out the original post here.

AI Alone Can’t Run Revenue

Finance doesn’t run on “mostly right.” It runs on math.

In The Architecture Behind AI-Native Revenue Automation, Tabs’s CTO breaks down why LLMs alone aren’t enough—and what it actually takes to build audit-ready, AI-driven contract-to-cash systems for modern B2B teams.

Get the whitepaper

*Ad