When should you use it?
- You pull the same web data on a regular cadence
- Manual copy-paste scraping wastes your time
- Your computer can be awake when the task runs
Not for: Scraping sites whose terms forbid it; Expecting scheduled browser tasks to run while the machine sleeps
What do you need, and what do you get?
You need
- A browser agent extension
- A page with the data you track
- A schedule for the recurring run
You get
- A ranking table exported as a CSV
- A reusable prompt saved as a recurring job
- A weekly dataset ready for analysis
How does it work, step by step?
- Write a precise scrape promptSpecify the exact columns and row count you want from the table.
- Export to CSVHave it return a downloadable file, not just text on screen.
- Save the working prompt as a taskTurn the prompt that worked into a recurring shortcut you can run later.
- Pick the cadenceSet it to run weekly, or whenever your cycle repeats.
- Know the limitsScheduled browser jobs only fire while the machine is on; use the desktop tool for file work.
The prompt
The original prompt by Frank Andrade, word for word, with a link to where it was published. Below it is our adapted version with fill-in blanks, if you want it. Works in Claude. Values in lime are placeholders.
Original prompt by Frank Andrade
Build a table with the following columns for the first 250 elements in the table: Ranking Newsletter Date (today’s date) Type = free/paid Give me a link to download the file in csv format
Exactly as published by Frank Andrade in How to Turn Claude into a 24/7 Employee (It Has Automated My Work).
Our adapted version (optional)
In my browser, scrape [page] into a table with these columns: [list columns] for the first [N] rows. Give me a downloadable CSV. Once this works, I will turn it into a recurring job that runs every [cadence]. Keep the extraction exact and do not summarize, I want the raw rows. Note anything on the page that could break the scrape next week.
What can go wrong?
- Browser tasks and desktop tasks are isolated; pick the right one for the job.
- Scheduled runs need the machine awake, so plan timing accordingly.
Our take
Frank Andrade's leverage comes from turning a prompt that already works into a scheduled job, which is where web scraping stops being a one-off demo. Be exact about the columns and the CSV layout so each week's file drops straight into analysis without cleanup. Scheduled browser tasks have practical limits: the machine must be awake, signed in and stable when the job runs, and they stay separate from desktop tools that can reach local files, so send file-heavy steps to the latter. Check the target site's terms and robots rules before putting a scrape on a timer, and prefer an official export or API if one exists. Cap how often it runs, keep a log of failures, and stop on unusual pages such as captchas rather than retrying. Treat the CSV as a working file you own, and compare each week's list with the last so sudden blank results are obvious. Congrats to climbing newsletters are fine; using the data for spam outreach is not.
Where the time goes back: A weekly scraping-and-formatting chore that takes ages by hand turns into a scheduled run that leaves a clean CSV waiting for you.
FAQ
How do I make it recurring?
What are the catches?
Source and credit
The original prompt on this page is by Frank Andrade, shown word for word with a link to where it was published. The summary, notes and adapted version are ours.
- Original
- How to Turn Claude into a 24/7 Employee (It Has Automated My Work)
- Creator
- Frank Andrade
- Published on
- artificialcorner.com
- Original date
- Mar 4, 2026
- License
- Not openly licensed; the original prompt is shown word for word with credit and a link to the original
- Originally for
- Claude
- AI dry run
- Oct 23, 2026: passed. Dry-run tested by our AI test agent on sample input. PASS: Gates held on desk check; no live sends or purchases.
AI dry run on sample data; not run with live connected accounts. The playbooks.bot test agent filled the prompt with synthetic inputs, produced the output and checked it against the page's stated outputs and guardrails. - Verified
- Oct 23, 2026: source link resolves to the original, the credited creator published it, our excerpt or prompt matches it, and the license allows our use.
Are you the creator? Claim or correct this page.