Hi Anna team,
I’ve been trialing Anna and I really like the core idea: a local agent that can actually execute tasks and work with my files, while the cloud side handles memory and settings sync. That part makes sense on paper.
In practice, though, the two-part experience feels a bit fragmented:
-
I have to download and install a desktop agent, but then I’m primarily interacting through a browser for the actual chat and day-to-day usage.
-
It creates constant context switching (app for status, web for conversation, web for settings), and it makes the product feel “split” rather than cohesive.
-
From a user trust perspective, “the thing that can touch my local files” being separate from “the thing I talk to” can feel less transparent, even if the architecture is technically sound.
To explain what I mean: tools like the Codex desktop app provide a single, unified experience where the chat UI, local file access, and project/workspace context all live in one place on the machine. You can still have account sync and cloud features, but the primary workflow doesn’t require living in a browser.
A few questions / suggestions:
-
Is there a roadmap for a fully integrated desktop app (chat + settings + agent) where the web dashboard is optional?
-
If the web chat is staying, could the desktop app embed the chat UI (or offer a “desktop chat” mode) so users don’t need to bounce between app and browser?
-
Are there specific constraints (security model, deployment speed, enterprise requirements) that made you choose “browser-first chat” instead of a single client?
-
Would you consider an “all-local” mode (with optional sign-in for sync) for users who want everything in one place?
I’m sharing this because the underlying functionality is strong, and a more unified UX would make it feel much more polished and intuitive. Curious to hear the reasoning behind the current design, and what direction you’re aiming for longer-term.
Thanks!