How We Moved Blog Image Uploads from Manual Server Work to the Admin UI and CDN

Publishing one article should not require a fresh code deployment or direct access to a server. Text, SEO fields, and images belong in the same editorial workflow—especially when a team intends to publish every day.

TekeraLab already had a working CDN. The missing part was the final connection between that permanent media layer and the authenticated Admin Blog interface.

The problem was not the absence of a CDN

A CDN can deliver files quickly and reliably, but it does not automatically give an editor a safe way to place those files there. Before the Admin workflow was completed, an image could still require someone with technical access to move it, find the public URL, and paste that address into the right field.

That may be manageable for an occasional article. At daily publishing volume, it becomes a bottleneck and gives infrastructure work to people who should be editing content.

The missing link between Admin and the existing CDN

The right solution was not another storage system. It was an authenticated upload path to the CDN we already operated. An authorized editor should be able to choose an image, specify how it will be used, upload it, and receive a stable public URL without leaving the article editor.

Cover, Open Graph, and body images can then be managed beside the title, article body, translations, and SEO metadata. A new article no longer needs code changes simply to make a picture public.

A secure upload is more than a button

The visible Upload button is only the front of the workflow. The important controls happen behind it: the endpoint must require an authenticated administrator and inspect the actual content of the file rather than trusting its filename.

File type, size, and destination need clear limits. A successful request should return one stable public URL, while a failed request should explain the problem without creating a half-finished media record. The goal is not accepting every file; it is providing one predictable gate for valid blog images.

The issue that only appeared when we looked at the real page

The API could accept a file and the CDN could return HTTP 200, yet the first real preview still exposed a problem: one URL form did not render correctly through the page’s image optimization path. Two technical components were healthy, but the reading experience was broken.

That is why visual verification matters. Upload success and a reachable file are necessary, but not sufficient. The image must appear inside the article on desktop and mobile, with its intended dimensions and alternative text, and without a broken frame.

What changed for everyday publishing

The desired path is now simple: the editor selects an image in Admin Blog, the file passes the permitted checks, the existing CDN returns a public address, and that address becomes the Cover, OG, or body image. The content team no longer needs extra server access.

This change also clarifies responsibility. Infrastructure maintains a safe and durable path; editors prepare and verify the article from the interface; and both groups can inspect the same public result. It is one of the practical foundations behind our four-language publishing workflow.

The missing piece was not a new CDN—it was the last metre of the publishing path. Connecting Admin Blog to the CDN turned image handling from manual server work into a normal editorial action. The most important lesson remained visual: a green API is not the end of the test; the reader must be able to see the image on the page.