Version main
CloudOS Direction
docs/CLOUD_OS_DIRECTION.mdx @ HEAD
CloudOS Direction (Archived Product Vision)
CloudOS is an AI-native, cloud-native operating system direction that treats the browser as the primary runtime and devices as capability providers.
Instead of targeting platform silos (Android, iOS, Linux, Windows), CloudOS abstracts OS-level differences behind device agents and cloud routing.
Vision
- Build an operating system layer above traditional OS boundaries.
- Let users interact through browser + devices, not platform-specific app stacks.
- Expose device capabilities through local backend agents.
- Keep user state synchronized across devices and cloud endpoints.
- Run AI-driven workflows across personal and remote infrastructure.
Core Model
-
Browser-first control plane: CloudOS UI runs in the browser and provides a unified command surface.
-
Device capability plane: Each device runs a backend agent that exposes capabilities (compute, files, camera, sensors, network reachability, local apps/services).
-
Workflow execution plane: Workflows can run on one or many devices and be routed through tunnels/overlay networks such as Cloudflare Tunnel, Tailscale, or other providers.
-
Integration plane: Cloud and SaaS integrations extend functionality for infrastructure and personal productivity domains such as email, contacts, calendar, and cloud storage.
Identity, Security, and Data Ownership
- Identity is anchored to a Solana wallet.
- User data is encrypted with wallet-linked cryptography.
- Storage is bring-your-own-medium:
- Google Drive
- GitHub
- Browser local storage
- Other user-controlled storage backends
- Login requires both:
- wallet authentication
- access to the selected storage medium containing encrypted profile/state
Resilience and Sync
- More connected devices increase redundancy and availability.
- Device agents replicate/sync encrypted state to reduce single points of failure.
- Users can recover state from any authorized storage backend they can access.
Economic Layer
- Users can opt in to accept workloads from other users.
- Contributed device capacity can be monetized.
- Scheduling/routing policies assign eligible jobs to participating devices under explicit constraints.
Status in This Repository
cloud-os/implementation is intentionally removed.- Direction is retained in this document and surfaced in
frontend/so the repository keeps a single frontend source of truth.