How I Turned One Prompt Engine Into 39 Sellable Products On A 2 GB Server

Description and use

The author built a pipeline that turns a single vector prompt engine into 39 sellable products, addressing metadata formatting and duplicate uploads, all on a 2 GB RAM server.

Problem

Metadata formatting mismatches and duplicate uploads caused failed bulk submissions to stock platforms, wasting time and risking account flags.

Outcome

The pipeline produced 39 sellable products from a single prompt engine, with 856 MB of finished output, while operating within strict 2 GB RAM constraints.

Workflow

• Read asset folder and generate platform‑specific CSV headers with make_metadata.py
• Use upload_tracker.py to track pending and completed uploads, preventing double‑posting
• Convert Markdown guides to styled PDFs with md2pdf.py
• Package assets, covers, and thumbnails into zips with pack_product.py
• Upload packaged products to stock platforms

Evidence

Adobe wants exactly five columns in a specific order. Vecteezy wants four. Dreamstime wants fifteen.

Source

The fix is dumb and it works: a ledger.

Source

The whole thing ran on a single-core box with 2 GB of RAM and no GPU.

Source