Anthropic develops managed 24/7 Agent for Claude mobile apps too

Anthropic appears to be preparing to bring managed agents (Conway) to its consumer-facing Claude apps, with evidence of the feature surfacing in the iOS version of Claude behind a hidden feature flag. The functionality, which would mirror what is already available on the web platform, suggests that Anthropic is working toward a unified experience across all its surfaces - desktop, web, and now mobile. The move follows Anthropic's recent release of managed agents on its developer workbench, where businesses can build containerized agents powered by Claude. Bringing a consumer version of this capability into the main Claude app would represent a natural extension of that infrastructure, potentially giving everyday users access to always-on agents that can operate within their existing Claude context, including memory, preferences, and connected tools. This would position managed agents not as a separate product but as a deeply integrated layer within the app users already know. > BREAKING 🚨: Anthropic is working on adding Claude Conway always-on agent to its mobile app. > > This means that we potentially may see a consumer version too! > > Soon? 👀 https://t.co/Ng1HSORQrB pic.twitter.com/gqmvGH7Vdm > > — TestingCatalog News 🗞 (@testingcatalog) April 7, 2026 Anthropic has established a pattern of rapid iteration, often shipping features with minimal lead time between internal testing and public availability. While no specific release date has been identified, the presence of feature flags in the production iOS build suggests the work is well underway rather than purely experimental. The company, which has been aggressively expanding Claude's capabilities across code, research, and productivity workflows, would be adding another dimension to its platform with persistent agent functionality. The broader competitive context makes this move unsurprising. Several Chinese AI companies like Kimi and MiniMax have already shipped their own versions of in-app managed agents, and standalone agent platforms like Manus have demonstrated strong user interest in autonomous, task-completing AI. Anthropic entering this space on mobile, where most consumer AI usage increasingly lives, signals an intent to compete not just on model quality but on the agentic experience layer that is rapidly becoming table stakes across the industry.