How to Make Claude Code Do Your Boring Tasks on Autopilot
Think about how many tasks you do every single week that follow the exact same pattern. Check your inbox for urgent emails. Prep for client meetings. Summarize what happened in Slack. Pull the latest industry news.
Now imagine waking up on Friday morning to an email that already lists every urgent message you haven't replied to, with links to each one. Or opening your laptop before a client call to find a full briefing document pulled from your email, Slack, Notion, and the web; already compiled and waiting.
That's the promise of Claude Code automations. And unlike most "automation" pitches, this one doesn't require you to be a developer, learn a new platform, or pay for another subscription.
AI educator Allie K. Miller recently walked through exactly how she set this up for herself, and it's one of the most practical Claude Code tutorials we've seen. We're going to break it down step by step, plus show you an even easier option if you're on a Mac. (Want more from Allie? Check out Allie's website here! She's amazing and creates amazing work in the AI space).
First up, the TL;DR
You know that feeling on Friday morning when you realize you forgot to reply to three urgent emails from Tuesday? Yeah, same.
AI educator Allie K. Miller got so fed up with this that she built a system inside Claude Code (Anthropic's command-line tool) that automatically scans her Gmail every Friday at 9 AM, finds every urgent email she hasn't responded to, and sends herself a summary with direct links to each thread.
No extra app. No subscription. Just a custom slash command and a schedule.
Here's the gist of how it works:
- Custom commands are reusable instructions saved in Claude Code. Miller's /ue command checks her inbox (read AND unread), flags urgent messages, and checks if she's already replied.
- MCP connections let Claude Code talk to your Gmail, Slack, Notion, or calendar. Think of it as giving Claude a set of keys to your work apps.
- Cron jobs run your commands on a schedule automatically; every morning, every Friday, whatever you want.
- Stacked commands are the power move. Miller built a daily briefing that checks her calendar, then automatically runs a separate client-prep command for every meeting that day. She wakes up to fully compiled briefings.
The whole setup takes 2-3 hours the first time (mostly getting Google permissions sorted). Miller says it saves that much time in a week or two.
Don't want to mess with cron jobs? runCLAUDErun is a free Mac app that lets you schedule Claude Code tasks with a visual interface instead of terminal commands. Same idea, friendlier packaging.
Why this matters: The "what should I actually use AI for?" question finally has a concrete answer for most knowledge workers. You're not replacing your job; you're automating the 30 minutes of information-gathering busywork that sits between you and the work that actually requires your brain. And once you set it up once, it runs forever. Well, as long as your laptop is open. Baby steps.
What Are Claude Code Commands?
If you've used Claude Code (Anthropic's command-line coding tool), you know you can type instructions and it'll do things on your computer. Commands take that a step further. They're reusable shortcuts you create once and trigger anytime.
Think of them like saved recipes. Instead of explaining the full dish every time, you just say "/carbonara" and it knows what to do.
Miller's example: she built a command called /ue (short for "urgent emails") that scans her Gmail for messages that need a response, checks whether she's already replied, and emails her a summary with links to each thread. One slash command, and she's got a prioritized to-do list for her inbox.
Method 1: Build It Yourself (The Allie Miller Walkthrough)
This is the more hands-on approach, but it's also the most powerful. Here's what Miller set up, broken into digestible steps.
Step 1: Create a Custom Command
Commands live as markdown files in Claude Code. You write out what you want Claude to do in plain English, give it a name, and it becomes a slash command you can trigger anytime.
Miller's urgent email command includes:
- What to search: All emails (read AND unread) from a configurable time window. She originally only checked unread emails, but realized she'd sometimes read an urgent email and forget to reply. Relatable.
- Response tracking: For each urgent email, the command pulls the full thread and checks if any replies came from her. It categorizes everything as "needs response" or "replied."
- Output format: It sends her an email with the results, organized with "needs response" items at the top, each linked directly to the email thread so she can click straight through.
Step 2: Add Arguments for Flexibility
Here's where it gets clever. Miller set up her command to accept arguments (i.e. modifiers you can tack on). Type /ue and it defaults to checking the last seven days. Type /ue 3 days or /ue 1 month and it adjusts automatically.
This means one command handles multiple scenarios. Checking in on a Tuesday night at 1 AM? /ue 1 day. Doing a monthly review? /ue 1 month.
Step 3: Connect Your Tools With MCP
This is the most involved step, but it's also what makes the whole thing work. MCP (Model Context Protocol) is how Claude Code talks to other apps. Think of it as giving Claude Code a set of keys to your Gmail, Slack, Google Calendar, or Notion.
Miller specifically used Anthropic's official Google Workspace MCP, which she recommends over third-party alternatives for security reasons. Even if a community-built option has more features, going with Anthropic's version gives you more confidence about what's accessing your data.
Fair warning: Miller says this step took about an hour her first time, between setting up OAuth permissions and troubleshooting. But you can literally paste errors back into Claude Code and ask it to help you fix them. Talk to it like you've never seen a computer before, she says. It'll walk you through every click.
Step 4: Schedule It With a Cron Job
A cron job (from "chronos," the root of time) is just a task that runs on a schedule. Instead of typing /ue every Friday morning, you tell your computer to run it automatically at 7 AM.
Miller set hers to run every Friday at 9 AM. Setting this up is as simple as telling Claude Code: "Set this command up to be on a cron job that kicks off every Friday at 9 AM."
A few things to keep in mind:
- Your computer has to be on. If your laptop is closed at 7 AM, the cron job won't run. Miller suggests either keeping your computer awake or getting a dedicated device (like a Mac Mini) that's always on.
- Authentication can expire. If your Google MCP token expires, the job fails silently. Miller recommends adding a fallback that emails you if something breaks.
- Test it first. Set a cron job for two minutes from now and make sure it actually works before you walk away.
You can verify your setup anytime by asking Claude Code: "Is it set up? Is it safe? Should I be nervous about this command?" It'll give you a safety assessment and help you catch potential issues.
Step 5 (Power Move): Stack Your Commands
This is where things get genuinely exciting. Miller describes "stacking," which means having one command trigger another.
Example: You create a /daily-brief command that checks your Google Calendar. It sees you have three client meetings today. For each client, it automatically spins off your /client-prep command, which pulls the last two weeks of emails, Slack messages, and web news related to that client.
You wake up. You open your laptop. Three client briefings are sitting in your inbox.
That's not hypothetical; that's what Miller built. The /client-prep command she demoed pulls from Gmail, Slack, Notion, and the open web, then compiles a briefing with key communications, active documents, and even suggested talking points based on recent pop culture news (for small talk).
Method 2: The Easy Button (runCLAUDErun)
Not everyone wants to wrangle cron jobs and MCP tokens. If you're on a Mac and want the simplest possible path, runCLAUDErun is a free macOS app that puts a visual interface on top of Claude Code scheduling.
Instead of writing cron syntax in a terminal, you get a clean native app where you can create tasks, pick a schedule (once, daily, weekly, or custom intervals), and let it run in the background. Everything runs locally on your machine; no cloud dependencies, no data leaving your computer.
What it handles for you:
- No cron syntax to learn. Point and click instead of memorizing "0 7 * * 5."
- Task history and logs. See what ran, when, and what happened.
- Automatic updates. New features install in the background.
You'll still need Claude Code installed and an active Claude subscription, but runCLAUDErun eliminates the terminal-heavy parts of the setup process. It's especially good if you want to start automating quickly and don't need the MCP integrations (yet).
What Should You Actually Automate?
Miller suggests giving Claude Code a detailed description of your job, your tools, and your daily workflow, then asking it to brainstorm commands that would save at least one hour per week.
When she demoed this for a hypothetical 50-person consulting company in supply chain, Claude Code suggested:
- Supply chain news: Automated daily research summaries.
- Competitor monitoring: Track what rival firms are doing.
- Tariff tracking: Especially relevant given current trade volatility.
- Client meeting prep: Pull context from email, Slack, Notion, and the web.
- Slack digest: Summarize key decisions, blockers, and wins across channels.
- Pre-drafted emails: Write personalized updates for every team member based on Slack activity.
The recurring theme: any task where you're gathering information from multiple sources, summarizing it, and acting on it is ripe for automation.
A Note on Learning vs. Outsourcing
Miller makes a point worth repeating. If you use Claude Code to just "do stuff" without understanding what it's doing, you won't learn anything; and you might create security problems you don't recognize.
Her suggestion: throughout the process, ask questions. "Did I set this up right?" "What's the smartest way to do this?" "Help me think through potential problems." If that means a task takes three hours instead of two, that's fine. Especially if you're non-technical, this is free education.
The Bottom Line
Setting up your first Claude Code automation (including the Google Workspace MCP) might take 2-3 hours. But Miller estimates it saves that much time in one to two weeks, and the savings compound from there.
- If you want maximum power and flexibility: Follow Miller's full tutorial for custom commands, MCP integrations, cron jobs, and stacked automations.
- If you want the fastest path to scheduling: Download runCLAUDErun (free, macOS only) and start with simple recurring tasks.
Either way, the era of manually doing the same thing every Monday morning is officially over. Your future self, the one who wakes up to a compiled inbox summary and three client briefings, will thank you.
P.S: If you're pretty technical and have a space comptuer lying around, you could handle these types of automations using OpenClaw. For more about that, read our article all about OpenClaw here.
Also, this is a snippet from a longer series we're writing on how to work with Claude Code to make your own automations. To read the whole thing, click here.
