Claude Code /statusline — A Must-Have When You’re Working Across Multiple Projects

If you’re running multiple AI coding agents in your terminal, you’ve probably had this moment:

“Wait… which project is this tab again?”

It sounds small, but it happens a lot. And it breaks your flow every time.

Here’s how I fixed it using Claude Code’s /statusline.

The Problem: Terminal Tabs All Look the Same

I use Ghostty as my main terminal. It’s fast, clean, and I really like it overall.

But there’s one thing missing:
you can’t name tabs manually.

Tools like SecureCRT let you label tabs as “frontend”, “user-service”, etc. That makes it easy to switch between projects.

With Ghostty, every tab just shows the same shell prompt. Once you start running multiple AI agents across different projects, everything blends together.

Codex Actually Solves This by Default

If you’ve tried OpenAI Codex CLI, you might have noticed something nice:

There’s always a small status line showing the current project path.

So even with multiple tabs open, you instantly know:

• this one is frontend
• that one is user-service

It’s a small detail, but surprisingly helpful.

Claude Code Has the Same Feature (It’s Just Hidden)

Claude Code supports this too — it’s just not enabled by default.

The feature is called /statusline.

How to Enable It

Just run this inside Claude Code:

/statusline show current directory, model, and context usage

That’s it.

Claude Code will automatically:

• create a script at ~/.claude/statusline-command.sh
• update ~/.claude/settings.json

No manual setup needed.

What You Get

After enabling it, you’ll see something like this at the bottom:

/home/jason/workspace/vita | Claude Opus 4.6 | Context: 12% used

This gives you:

• Current directory → which project this tab belongs to
• Model name → Opus, Sonnet, etc.
• Context usage → how full your conversation is

Now when you switch tabs, you don’t have to guess anymore. Just glance at the bottom.

Why This Actually Matters

In real workflows, you’re rarely using just one agent.

For example:

• Tab 1 → services/user (backend API work, Claude Code)
• Tab 2 → frontend (UI work, Claude Code)
• Tab 3 → services/chat (debugging, Codex)

Without context, you end up doing things like:

• running pwd
• scrolling up to find where you are
• double-checking before making changes

It’s small friction, but it adds up.

/statusline removes that completely.

The Underrated Part: Context Usage

The context indicator is more useful than it looks.

Claude Code compresses older messages as the conversation grows. When usage gets high, you might start losing useful context.

With /statusline, you can see it in real time and decide:

• continue the session
• or start fresh before things get messy

You Can Customize It

The best part? It understands natural language.

You can ask for exactly what you want:

/statusline show git branch, project name, and cost

or

/statusline show directory, model, context percentage, and session cost in USD

Common options include:

• current directory
• model name
• context usage
• git branch
• session cost

Mix and match however you like.

Final Thoughts

If you’re working with multiple AI agents across different projects, this is one of those tiny features that makes a real difference.

It takes about 30 seconds to set up, and you’ll feel the benefit immediately.