Anthropic Launches Claude Reflection Dashboard in Beta
SkimNews Take
Reflect" tasks users with auditing their own AI dependency, but only when memory is enabled, and it deliberately excludes connected-tool usage — so the most consequential delegation channels stay outside the mirror's view.
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Anthropic rolled out "reflect" on Thursday — a Claude reflection dashboard summarizing users' key topics, delegated task types, and peak usage times across month-to-year windows, inspired by the Spotify Wrapped format.
- The reflection dashboard also lets users set custom "quiet hours" or break reminders after extended sessions, with a total-time-spent metric "coming soon."
- Anthropic's prompt design includes reflective questions like "What's one thing you want to keep doing yourself, even if Claude could do it faster?" — after which users can talk their answers through with Claude itself.
- The privacy design excludes files inside connected platforms (e.g., actual emails in a linked inbox), incognito chats, and health-integration conversations, surfacing "sensitive topics" only at a high level.
- Reflect is in beta for free, Pro, and Max subscribers via Settings on web and desktop, requires Claude memory to be on, and was built from interviews with Claude users.
- The feature will come to Claude Cowork "soon," extending the self-reflection framing into Anthropic's collaborative product.
Why it matters: Anthropic is building brand identity around the "AI collaborator" rather than AI doer, routing every reflection back to "talk it through with Claude" — a striking circular design. With the feature expanding to Claude Cowork, that self-reflection framing moves from individual chats into team workflows, while Anthropic pre-empts pushback by carving out privacy boundaries (no connected-tool file contents, no health data) before launch.
Ask SkimNews
