The post editor's quality check moved into the document itself. Flagged claims get a wavy underline; clicking one shows the reason, an AI-suggested fix grounded in your own pages, or the option to edit or keep it. Publish and Schedule stay off in the toolbar while a claim marked possibly false or invented is unresolved, or while the check hasn't seen your latest edits. This same gate applies on the Posts dashboard, bulk actions, the worksheet, and agent and MCP tools — each refusal names the held post and links back to the editor.
Visitors reading a brand page can now click "Compare your brand" to sign up, run their own scan, and land in results that compare against that brand from day one. Brand owners can also claim their page and re-run their scan. When the login email matches the brand's domain, their scan replaces ours on the public page and the page shows "Verified by ".
Social media posting is gone entirely. Instagram, X, LinkedIn, and TikTok creation, scheduling, and publishing have been removed, along with the Social history page and the "Create Marketing Posts" action on blog posts. Blog posts are unaffected — publishing to your agent site and syncing to WordPress, Webflow, or Ghost all work as before.
Brand pages without a real scan no longer appear in search indexes or the brands sitemap. Blog post meta descriptions now render correctly when they contain apostrophes or quotes — previously, a description like "Motion graphics don't stop…" was served as "Motion graphics don". AI agent visits to brand subpaths now count correctly in Site Analytics instead of reading as zero.