The Atlas Roadmap
In Progress
Phase 0 — Trust-First Foundation
Atlas is designed to be safe before it is smart.
- Deterministic intent handling
- Explicit confirmation before changes
- Offline-first and private by default
- Clear failure explanations
Planned
Phase 1 — Daily Personal Utility
Reduce cognitive load without removing agency.
- Natural language recall
- Cross-hub awareness
- Preparation and reminder prompts
Future
Phase 2 — Learning Without Risk
Learning that suggests, never decides.
- Pattern detection
- Transparent reasoning
- User-controlled learning
Roadmap Disclaimer
This roadmap reflects our current direction, not a delivery guarantee.
Features may change as we prioritise safety, ethics, and user trust.