S2-B1: Native Presence

Intent

Give Seraph a native-feeling local presence so it no longer depends on a browser tab and manual startup habits.

Capabilities in scope

• Tauri desktop shell or equivalent native wrapper
• tray presence
• system notifications
• startup and background lifecycle handling
• packaging that feels closer to an app than a dev environment

Non-goals

• full cross-platform parity on day one
• complete replacement of the Python backend
• deep mobile support

Required architectural changes

• define the desktop shell boundary and process model
• unify lifecycle management for backend, daemon, and UI
• add notification and tray interfaces
• harden local packaging and startup paths

Likely files/systems touched

• desktop packaging layer
• manage/run lifecycle assumptions
• notification and status surfaces
• setup and docs for installation

Acceptance criteria

• Seraph can be launched and kept present as a local app
• notifications work without relying on an open browser tab
• local lifecycle is simpler than the current Docker-first path
• desktop presence does not break existing web development flow

Dependencies on earlier batches

• depends on Season 1 making the runtime worth packaging

Open risks

• packaging complexity can slow momentum
• desktop shell work can distract from core product behavior if not tightly scoped
• process orchestration may become brittle without good observability