Split homework scans by student and update the roster

A Claude Code skill sorted scanned homework pages by handwritten name, created per-student PDFs, and marked completion in a spreadsheet.

Done with Claude Code· Intermediate

What they did
They set up a Claude Code skill to take a combined PDF of scanned homework pages. The skill used vision to read each handwritten student name, fuzzy-matched it against a roster, then separated the file into one PDF per student and filled in an Excel completion sheet. It also supported session resume for batches larger than 99 pages.
How it went
The setup was meant to keep student attribution accurate through a manual check step. It also helped identify missing assignments across large class rosters. No timing or output counts were reported.
Tips
Keep a human in the loop to verify name matches before finalizing anything. For very large batches, make sure the workflow can resume after 99 pages.

Try it yourself with Claude Code

Build a Claude Code skill for [folder path] that reviews a batch of scanned homework PDFs, reads the handwritten name on each page, matches each name to [roster file] with fuzzy matching, and groups pages into one PDF per student. Update [completion spreadsheet] to show who submitted work, and produce a missing-work list plus any uncertain matches. Before you write or change files, show me the planned file splits and spreadsheet edits for approval; if I approve, proceed, but pause again before any irreversible action such as deleting originals, sending files, or submitting changes. If the batch has more than 99 pages, continue by resuming the session rather than stopping.

Fill in the [brackets]. Agents can make mistakes, so review anything before it is sent, paid or deleted.

Read the original on mcpmarket.com ↗

Get 3 agent uses like this every morning

With the day's top AI news. Free, and one click to unsubscribe.