Custom Commands

Module 6: Custom Commands

Every workflow you run more than once is a candidate for a command.

By this point you’ve got habits: the morning check-in, inbox processing, the weekly review. These are good patterns. But right now each one requires you to type a prompt from scratch, and small variations in how you phrase things lead to different results.

Custom commands fix that. You define the exact prompt once, trigger it with a slash command, and get consistent results every time.

Why This Matters

Think about the difference between asking a colleague “can you check my notes and see what I need to do?” versus handing them a structured checklist: “Here’s what to look at, here’s how to prioritise, here’s the format I want the output in.”

The second version produces better work. Every time.

Custom commands let you encode the second version — the careful, specific prompt — and trigger it in two keystrokes. Over time, you build a library of these. The library is your personal toolkit, and it gets more valuable the longer you use it.

Anatomy of a Slash Command

In Claude Code, slash commands live in markdown files inside a .claude/commands/ folder in your vault root. Create that folder now:

mkdir -p ~/Vault/.claude/commands

Each command is a markdown file with a name that becomes the slash command. Create your first one:

~/Vault/.claude/commands/daily-review.md

Review my vault for today's priorities.

1. Check 02 Projects for any notes modified or created today
2. Check 01 Inbox for unprocessed items
3. Look at today's daily note if it exists
4. Summarise: what's active, what's waiting, what needs attention

Format the output as a brief list. Keep it under 200 words.

Save the file. Back in Claude Code, type /daily-review and press Enter. Claude reads that file and executes the instructions exactly.

That’s it. One file, one command, repeatable every day.

Building Your Command Library

Start with the workflows you already have. Look at what you’ve been typing repeatedly and turn each one into a command.

Good candidates for early commands:

/morning — your morning check-in. Review inbox, projects, calendar note. Tell me what to focus on.

/process-inbox — read each item in 01 Inbox, suggest where it belongs, and list items I need to action rather than file.

/weekly-review — run across the week’s daily notes, completed project updates, and any captures. Give me a summary of what happened and what’s carrying forward.

/capture — take what I’m about to type and create an inbox note with today’s date and the right metadata.

Each command you write encodes your judgment about what a good version of that workflow looks like. You’re not just automating a task — you’re capturing your best thinking about how to do that task.

When to Create a Command

Create a command when:

• You’ve typed a similar prompt three or more times
• The output needs to be in a specific format every time
• You want to share the workflow with a future version of yourself (or someone else)

Don’t create a command for things you do once or things that change significantly each time. Keep the library for genuine repeatable patterns.

Before moving on:

• .claude/commands/ folder created in your vault
• At least one slash command written and tested
• The command produces the output you wanted

---

This module evolves as your system does.