# Session 05 — Materials

This folder has the starter data for Day 2, Session 5.

## Files

| File | Purpose |
|---|---|
| `body.json` | The curl request body used in Part A. Saving the body to a file lets the same `curl -d @body.json` command work on macOS, Linux, Windows `cmd`, PowerShell, and Git Bash without modification. |
| `sample-sentences.txt` | A 10-line subset of the session 04 corpus. Use it to test your HTML LLM tool: one prompt per line. |
| `nlvnpf-0100-001.jpg` | A page from a Vietnamese manuscript in classical Han script. Used as the OCR test image in Part E (Stage 3 of the tool). Source: Nôm Foundation / National Library of Vietnam digitization. |

## In session

You will **not** write any of the HTML tool by hand — Codex builds it from a single detailed prompt in Part C, then extends it in Part D. This folder just gives you a small file you can upload through the tool to test that it works end-to-end.

Once your tool is working, swap in your own newspaper sentences, research notes, archive transcriptions, or whatever else you want batch-processed.