How pavrus117/ai-os-maps-guide works
This repository is a written setup guide, not an implementation: README.md provides interview and build prompts for a four-layer personal AI operating system, and ai-os-maps-guide.pdf is its companion copy. The principal workflow is to interview the user, establish a navigable memory map, share it across machines, run scheduled routines, then view or trigger work from a dashboard. The graph captures documented responsibilities and relationships, not verified software calls; the README excerpt ends during Card 3, so detailed Pulse and Screen wiring is limited to what it states.
Main entry points: README.md Always-on workspace
Components 21
Memory
- Interview session
README.md - Interview record
README.md - Master signpost
README.md - Area signposts
README.md - Work documents
README.md - Map checker
README.md
Agent
- Laptop workspace
README.md - Always-on workspace
README.md - Git synchronization
README.md - Private network
README.md - Restricted profile
README.md - Telegram bot
README.md
Pulse
- Routine list
README.md - Schedulers
README.md - Scheduled routine
README.md - Run records
README.md
Screen
- Dashboard
README.md - Selected machine
README.md
Other components
- User
- Claude Code
- Telegram
Connections 25
- User → Interview session: answers questions
- Interview session → Claude Code: guides interview
- Claude Code → Interview record: records answers
- Interview record → Master signpost: informs draft
- Claude Code → Master signpost: drafts map
- Master signpost → Area signposts: points to
- Area signposts → Work documents: points to
- Map checker → Master signpost: checks paths
- Map checker → Area signposts: checks sections
- Map checker → Work documents: flags omissions
- Laptop workspace → Git synchronization: shares workspace
- Always-on workspace → Git synchronization: shares workspace
- Git synchronization → Laptop workspace: syncs changes
- Git synchronization → Always-on workspace: syncs changes
- Private network → Always-on workspace: protects access
- Restricted profile → Always-on workspace: limits tools
- Telegram → Telegram bot: sends messages
- Telegram bot → Restricted profile: uses profile
- Telegram bot → Always-on workspace: logs replies
- Routine list → Schedulers: defines schedules
- Schedulers → Scheduled routine: runs when due
- Scheduled routine → Run records: records run
- Dashboard → Run records: reads records
- Dashboard → Work documents: reads data
- Dashboard → Selected machine: dispatches work