HO14 — Coded Workflow Automation

Hands-on portfolio project · Week 6 · Solo · module M18. Part of the ForgeMind AI — AI Productivity Essentials course.

Goal

Done when: A scheduled, coded automation (Claude Code) · measured time saved

What to ship

The automation script + how it's scheduled + a 5-step Loom + a measured before/after time log.

Pick a problem statement

Choose one of these real use-cases — or bring your own (get it approved first):

1. You copy the same numbers off a website into a tracking sheet every morning and it is soul-crushing. Build a coded automation with Claude Code that scrapes that page daily, cleans the data, and appends it to a sheet or database, scheduled on a cron. Success: it runs unattended each day and your time log shows the daily minutes saved.

2. You manually pull from an API, reshape it, and drop it somewhere every day. Build a coded pipeline with Claude Code that fetches from the API, transforms the data, and stores it or sends a notification, triggered on a schedule. Success: a cron-driven run that completes the fetch-transform-store loop unattended, with a before/after time log.

3. Files land in a folder and someone has to process each one by hand. Build a file-watcher automation with Claude Code that detects a new file, processes it (convert or extract), and writes the output. Success: dropping a file triggers the job automatically and produces the result, with a time log showing the manual minutes removed.

4. Your boss expects the same report emailed every week and assembling it eats an hour. Build a scheduled job with Claude Code that queries the data, builds the report, and emails it weekly. Success: the email arrives on schedule with no manual steps, and a before/after time log proves the recurring hour saved.

5. Two of your systems drift out of sync and you reconcile them by hand. Build a coded sync job with Claude Code that keeps a sheet, database or tool aligned automatically on a schedule. Success: the systems stay in sync after each scheduled run with no manual reconciliation, documented with a before/after time log.

How to use this repo

1. Click Use this template to create your own copy.
2. Build your chosen project in your copy.
3. Replace this section of the README with: what you built, the problem it solves, and how to run it.

---

HO14 · Solo · ForgeMind AI Course · module M18 (Week 6)

💡 Use your Claude.ai Pro plan wisely

The Pro plan has a usage limit that resets every few hours. A few habits make it stretch — and keep a mistake from burning your whole session:

• Use the example prompt in each sample's README — it's already written and tested. Don't reinvent it.
• One clear prompt beats lots of vague back-and-forth. Say what you want, with an example, in a single message.
• Start a new chat when you switch tasks. Long chats re-read every earlier message and use up your limit faster.
• Don't paste big files over and over. Paste once, then refer back to it.
• If something works, keep it. Tweak it rather than regenerate from scratch.
• Using Claude Code or Cowork? This repo's CLAUDE.md makes Claude follow these same rules automatically, and SKILL.md is a reusable "token-wise" skill.

If you do hit the limit, it resets after a few hours — nothing you've saved is lost.

Run it locally

This is a scheduled, coded automation — it runs on your own machine, not on the GitHub Pages page (Pages is static and can't run a cron job). To run and schedule it:

# clone your copy of this repo, then from your project folder:
pip install -r requirements.txt

# run it once to make sure it works:
python main.py

# then schedule it with cron so it runs unattended.
# edit your crontab:
crontab -e

# example — run every day at 8am:
0 8 * * * cd /path/to/your/project && /usr/bin/python3 main.py >> run.log 2>&1

On Windows, use Task Scheduler instead of cron to run the same command on a schedule. Each samples/sample-0X/ folder has its own README with the exact run command and a suggested schedule. Keep a before/after time log so you can measure the minutes saved.

Deploying to a server

You don't have to host this anywhere to pass — it runs locally on your own schedule. If you later want it running unattended on a public server (so it fires even when your laptop is off), the course covers that once: See the Week-6 deploy walkthrough.