ByHuman

Clean up X with local, reversible feed filters. Optionally create writing-process receipts on X and Gmail.

Make X readable again. ByHuman changes posts matching your enabled crypto-spam promotion and excessive formatting filters. Every change explains why and can be restored. Not an AI detector. Feed Filter Dim, collapse, or hide posts with observable low-quality signals: crypto spam, promo blasts, engagement bait, generic advice, formulaic threads, and hashtag walls. You choose the categories and the action. Every decision is explained, reversible, and correctable. Classification runs entirely in your browser by default. An optional AI engine (opt-in, requires sign-in, 500 free analyses per month) sends post text to our server for higher-accuracy classification. Only a hash is cached. Nothing is stored or used for training. WRITING RECEIPTS Write a tweet or an email as usual. ByHuman tracks how the text was composed, including what was typed, pasted, and deleted, right in the composer, locally. When you post, you get a tamper-evident receipt: a signed proof page showing your writing process. Add an optional stamp line so readers can verify it in one click. Your text is never stored. Only a cryptographic hash is stored. PRIVACY No browsing history. No reading outside the active composer and the visible feed. Full policy: https://byhuman.ink/privacy ByHuman does not detect AI authorship because no tool can reliably do so. It proves observable process for writers who want to show their work and filters observable patterns for readers who want a cleaner feed.