About
Nashra MENA is an aggregator. We collect headlines, short snippets, publication times, and OG images from a curated list of news sites across the Middle East and North Africa. Every headline links back to its original publisher. We never rehost article bodies, never quote more than 280 characters per item, and never run our own reporting.
Our source list is editorial. We picked outlets with the goal of representing the press of each country we cover — state media, opposition press, independent press, business press, and pan-Arab wires. The full list is visible in the sidebar by nation. Adding or removing a source is a manual decision; we publish the rationale here when notable changes land.
What ‘curated’ does not mean: we do not rank stories editorially, we do not write headlines, we do not summarize across sources, and we do not omit stories from a source we've chosen to carry.
Earlier versions of Nashra labeled each source by political alignment. We removed those labels in May 2026 (DECISIONS D-049): at MENA scale, a coherent multi-country alignment vocabulary is a curation problem we are not staffed to solve, and pretending otherwise was less honest than omitting labels and showing nationality plainly.
What you see now is the source's name, country, language, and the actual headline they ran. When the same story runs in multiple papers, we stack them so you can compare framings side by side. Drawing conclusions about alignment is left to you.
Nashra MENA covers the Arab/MENA press. We deliberately do not aggregate Israeli outlets, even when they cover regional stories. This is an editorial choice, not a technical limitation. If you want Israeli press coverage of the same stories, an English-language equivalent exists at haaretz.com or timesofisrael.com.
We also do not aggregate from social platforms (Twitter/X, Telegram, Reddit), pirate media, or paywalled outlets we can't link to in good faith. If a source disappears, paywalls fully, or starts blocking our scraper, we drop it from the list.
When we ingest an article we compute a textual signature (MinHash on the headline + first 200 characters of the snippet, with Arabic and Latin script normalized so paraphrases bucket together). Articles whose signatures fall in the same bucket within a 24-hour window form a cluster. The header of each card shows N sources and the country flags that participate. We never write a synthesized ‘neutral’ headline — the cards stack the publishers' actual headlines as they ran.
We use a one-time geo lookup on your first visit to send you to the country page that matches your origin. The lookup comes from a request header set by our edge runtime; we do not store your IP, do not run client-side analytics, do not set advertising cookies, and do not embed third-party scripts. Your country selection is stored in your browser's localStorage.
We will get sourcing wrong sometimes. A source disappears, a feed corrupts, a clustering algorithm folds two unrelated stories into one. When we notice, we fix it. When you notice first, the same fix is welcome — we just don't publish a contact channel right now (see DECISIONS §17 for context). Open a PR if you find this codebase.