every month as a senior designer at modern practice, i have to source blog images for our 40+ monthly clients. it's a time-consuming task, especially since each client needs a unique image that represents their monthly blog, sized according to their website design, and the image must be optimized for speed.

i vibe-coded a local app that does the work for me! i used apis from google sheets, pexels, and tinypng to create a blog image generator. it outputs 3 optimized images in the correct aspect ratio depending on the client.

it took me roughly 1 hour to code this application using chatgpt.

At Modern Practice, we often design 30+ page websites for our healthcare clients. In such fast-paced builds, copy, design, QA, and development handoffs can get messy, and small text mismatches between the approved master copy doc and the live Figma file often slip through.

To solve this, I vibe-coded a native Figma app powered by the Claude API that automatically cross-checks pasted copy against the selected Figma page. It highlights any inconsistencies between the two, ensuring the copy is always accurate before handoff.

This simple tool saves hours of manual QA and prevents client feedback loops. It took me just a few evenings to prototype.