Users can now see why X hides their posts from the For You timeline
X released ranking code and an "Under the Hood" tool showing labels that limit post visibility, including signs of shadowbanning.
X has open-sourced more of the code behind its For You timeline and launched a new tool that lets users see whether labels applied to their account or posts are limiting their visibility. The announcement came from the official @XOpenSource account on August 13, 2026.
The release expands earlier transparency work, building on the March 2023 decision to publish large parts of the recommendation algorithm on GitHub and a September 2025 announcement from X's engineering team. The 2026 release adds model configuration, filter details, and core ranking system information to what is publicly available.
The new "Under the Hood" page sits in the app's settings and is available to eligible users in a randomized pilot. Eligibility requires an account to be at least one year old, and users need 10 or more posts in the prior month for visibility-label data to appear. The tool shows account-level and post-level labels applied over the past calendar month, covering categories such as spam and adult content, and lets users download that data as a JSON file.
X Open Source framed the move as a direct response to long-running questions about hidden reach limits. "We're releasing the code that determines how posts are ranked and filtered in the For You timeline, including the content labels that can impact post visibility," the account wrote. "This builds on our previous open-source work, adding model configuration, filter and core ranking system details."
The code shows how the For You feed blends in-network posts from accounts a user follows with out-of-network posts from accounts they do not follow, running both through shared filtering, scoring, and truncation pipelines. Ranking relies on predicted interactions: likes and replies add positive weight, while reports carry a heavy negative score that can severely limit reach. One analysis of the open-source code cited by ExplainX notes that a reply can add roughly 5 to 20 points and a like about 0.5 points, while a report can subtract 234 points.
The move matters for Nigerian and African publishers, journalists, and creators who depend on the For You feed for distribution. Open access to ranking weights and visibility labels gives regional users a concrete way to audit whether content from the continent is treated fairly, rather than relying on anecdotal reports of shadowbanning. African data scientists and researchers can now study the recommendation pipeline directly, and civil society groups tracking election-related content gain a public benchmark for understanding how platform rules are applied.
Coverage from TechCrunch and Business Standard framed the change as a way for users to see whether they have effectively been shadowbanned. No regulatory directive is tied to the release; X presents it as a corporate transparency and open-source initiative. The feature is global, with no Africa-specific or Nigeria-specific implementation details in the announcement.
