All micro-internships
Micro-internship brief ≈16h across 4+ days — build on Day 1, let it run 3+ daysbuilder

Build a robot that works while you sleep (GitHub Actions)

The best engineers automate themselves out of chores. Use AI to build a scheduled GitHub Actions workflow that does something useful every day without you — fetch and commit a data snapshot, update a README stat — and show at least 3 real scheduled runs.

AI Application DevelopmentWorkflow Automation

Posted by The AI Internships

The work brief

  1. 01Pick a daily job: snapshot your city’s AQI/weather/prices into the repo, auto-generate a daily study checklist issue, update a "days until exam" badge.
  2. 02Write the workflow: a cron-scheduled GitHub Action running your script (AI-assisted). Get it running on demand first, then on schedule.
  3. 03Let it run for 3+ days. Scheduled runs WILL surprise you (timezone confusion is UTC’s favourite prank) — log what happened.
  4. 04Document in the README: what it does, when it runs, and the run history.

What you’ll produce

4 deliverables

Submission standard

Submit the repo, a screenshot of the Actions history showing 3+ scheduled (not manual) runs, and your explanation + surprises log. Include your AI workflow: tools, best prompts, and what you changed from the AI’s output. Your Actions screenshot must show the 'schedule' trigger on 3+ different dates, and the repo should contain the dated automated commits those runs produced.

  • Public repo with the workflow and its output

    Public repo with the workflow and its output — the commit history should include the dated automated commits your robot made on different days. A repo with no dated run outputs fails the authenticity gate.

    Public linkRequired

  • Screenshot of Actions history: 3+ runs triggered by schedule

    Screenshot of the Actions run history showing 3+ runs with the 'schedule' trigger (not 'workflow_dispatch'/manual) on 3+ different dates. Manual-only or same-day runs fail the authenticity gate.

    Image uploadRequired

  • What your robot does daily, and why that’s worth automating

    Written responseRequired

  • What surprised you (timezones, permissions, failures) and how you fixed it

    Written responseRequired

You’ll complete these inside your private workspace.

How your work is evaluated

The passing benchmark is 70/100.

Really scheduled

29%

3+ runs genuinely triggered by cron (not manual), visible in history and repo output.

Useful job

14%

The daily output has real value, visible in the repo.

Understanding

29%

The surprises log shows real debugging of scheduling/permissions.

Ran on a schedule over days

29%

Fails if the runs are manual/same-day, the screenshot lacks the schedule trigger and distinct dates, or the repo shows no dated automated commits from the runs.

How we grade your AI usage

30% of your score

Using AI is the point — it’s the skill this certificate proves. You’ll answer three short questions about how you used it: what you asked, what was wrong with its first answer, and what you changed. Specific, honest answers score high. “I pasted the brief and submitted the answer” scores near zero.